U.S. patent application number 16/081878 was filed with the patent office on 2021-07-15 for channel state information feedback method and device.
This patent application is currently assigned to CHINA ACADEMY OF TELECOMMUNICATIONS TECHNOLOGY. The applicant listed for this patent is CHINA ACADEMY OF TELECOMMUNICATIONS TECHNOLOGY. Invention is credited to Runhua CHEN, Wenhong CHEN, Quibin GAO, Hui LI, Rakesh TAMRAKAR.
Application Number | 20210218456 16/081878 |
Document ID | / |
Family ID | 1000005496044 |
Filed Date | 2021-07-15 |
United States Patent
Application |
20210218456 |
Kind Code |
A1 |
LI; Hui ; et al. |
July 15, 2021 |
CHANNEL STATE INFORMATION FEEDBACK METHOD AND DEVICE
Abstract
A CSI feedback method and a CSI feedback device are provided.
The CSI feedback method includes: determining a first precoding
matrix in a set of first precoding matrices, and determining a
first precoding matrix index value corresponding to the first
precoding matrix; determining a second precoding matrix in a set of
second precoding matrices, and determining a second precoding
matrix index value corresponding to the second precoding matrix,
the second precoding matrix including a beam selection section for
selecting beams from the beam groups for each antenna group and a
phase adjustment section for adjusting phases among the plurality
of antenna groups; and feeding back the first precoding matrix
index value and the second precoding matrix index value to an
access device, so as to enable the access device to select
corresponding precoding matrices from the set of first precoding
matrices and the set of second precoding matrix in accordance with
the first precoding matrix index value and the second precoding
matrix index value, and implement calculation on the selected
precoding matrices to obtain a precoding matrix for data
transmission.
Inventors: |
LI; Hui; (Beijing, CN)
; GAO; Quibin; (Beijing, CN) ; CHEN; Runhua;
(Beijing, CN) ; CHEN; Wenhong; (Beijing, CN)
; TAMRAKAR; Rakesh; (Beijing, CN) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
CHINA ACADEMY OF TELECOMMUNICATIONS TECHNOLOGY |
Beijing |
|
CN |
|
|
Assignee: |
CHINA ACADEMY OF TELECOMMUNICATIONS
TECHNOLOGY
Beijing
CN
|
Family ID: |
1000005496044 |
Appl. No.: |
16/081878 |
Filed: |
February 21, 2017 |
PCT Filed: |
February 21, 2017 |
PCT NO: |
PCT/CN2017/074248 |
371 Date: |
August 31, 2018 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04B 7/0486 20130101;
H04B 7/0617 20130101; H04B 7/0634 20130101; H04B 7/0639
20130101 |
International
Class: |
H04B 7/06 20060101
H04B007/06; H04B 7/0456 20060101 H04B007/0456 |
Foreign Application Data
Date |
Code |
Application Number |
Mar 11, 2016 |
CN |
2016101407122.2 |
Claims
1. A Channel State Information (CSI) feedback method, comprising:
determining a first precoding matrix W.sub.1 in a set of first
precoding matrices, and determining a first precoding matrix index
value corresponding to the first precoding matrix W.sub.1, the
first precoding matrix W.sub.1 consisting of beam groups for a
plurality of antenna groups of an antenna array; determining a
second precoding matrix W.sub.2 in a set of second precoding
matrices, and determining a second precoding matrix index value
corresponding to the second precoding matrix W.sub.2, the second
precoding matrix W.sub.2 comprising a beam selection section for
selecting beams from the beam groups for each antenna group and a
phase adjustment section for adjusting phases among the plurality
of antenna groups; and feeding back the first precoding matrix
index value and the second precoding matrix index value to an
access device, selecting, by the access device, corresponding
precoding matrices from the set of first precoding matrices and the
set of second precoding matrix in accordance with the first
precoding matrix index value and the second precoding matrix index
value, and implementing calculation on the selected precoding
matrices to obtain a precoding matrix W for data transmission.
2. The CSI feedback method according to claim 1, wherein the
precoding matrix W is calculated using the following equation: W =
W 1 .times. W 2 = [ X 0 0 X 1 X 2 0 X M - 1 ] W 2 ; ##EQU00082##
the first precoding matrix W.sub.1 is a block diagonal matrix
comprising elements of 0 other than diagonal blocks where a beam
group matrix X.sub.i is located, i=0, 1, . . . , M-1, M represents
the quantity of the antenna groups for the antenna array, X.sub.i
represents a beam group for an i.sup.th antenna group and it is a
N.sub.i.times.L.sub.i matrix with each column representing one beam
direction, N.sub.i represents the quantity of antenna elements in
the i.sup.th antenna group, N represents a total quantity of
antenna elements for the antenna array and N = i = 0 M - 1 .times.
N i , ##EQU00083## and L.sub.i represents the quantity of beams in
the beam group for the i.sup.th antenna group; the second precoding
matrix W.sub.2 is used for beam selection and phase adjustment, and
it is capable of selecting several beams from each beam group and
adjusting phases among the selected beams for each antenna group,
to acquire a final precoding matrix; and the second precoding
matrix W.sub.2 is a ( i = 0 M - 1 .times. L i ) .times. r
##EQU00084## matrix, where r represents a rank of the precoding
matrix W.
3. The CSI feedback method according to claim 2, wherein W 2 ' = [
E 0 0 E 1 E 2 0 E M - 1 ] W p , .times. where .times. [ E 0 0 E 1 E
2 0 E M - 1 ] ##EQU00085## represents the beam selection section
and is to select r.sub.1 beams from each antenna group, a beam
selection vector for the i.sup.th antenna group is defined as
E.sub.i=.left brkt-bot.e.sub.i,0 e.sub.i,1 . . .
e.sub.i,r.sub.1.sub.-1.right brkt-bot., where e.sub.i,j represents
a column selection vector having a length of L.sub.i and it is to
select one beam from the beam group matrix X.sub.i or merge the
beams in X.sub.i into one beam, j=0, 1, . . . r.sub.1-1, W.sub.p
represents the phase adjustment section, and r columns of W.sub.2'
are selected as the second precoding matrix W.sub.2.
4. The CSI feedback method according to claim 3, wherein when the
r.sub.1 beams are orthogonal to each other, W.sub.p=UI.sub.r.sub.1,
where U represents an orthogonal matrix comprising r.sub.2 columns,
each column comprises M phase adjustment factors, r.sub.2 is a
minimum integer greater than or equal to r r 1 , ##EQU00086##
I.sub.r.sub.1 represents a r.sub.1.times.r.sub.1 unit matrix; and
when the r.sub.1 beams are not orthogonal to each other, every
r.sub.1 groups of M antenna groups form M r 1 ##EQU00087## superior
groups, and W.sub.p=UV, where U represents an orthogonal matrix for
adjusting phases among the superior groups and comprising r.sub.2
columns, each column comprises M r 1 ##EQU00088## phase adjustment
factors, r.sub.2 is a minimum integer greater than or equal to r r
1 , ##EQU00089## V is used for adjusting phases among the antenna
groups in one superior group and V=.left
brkt-bot.V'(:,1)e.sub.r.sub.1.sup.(1) V'(:,2)e.sub.r.sub.1.sup.(2)
. . . V'(:,r.sub.1)e.sub.r.sub.1.sup.(r.sup.1.sup.).right
brkt-bot., V' represents an orthogonal matrix comprising r.sub.1
columns, each column comprises r.sub.1 phase adjustment factors,
V'(:,i) represents an i.sup.th column of V', and
e.sub.r.sub.1.sup.(i) represents an i.sup.th column of a unit
matrix I.sub.r.sub.1.sub..times.r.sub.1.
5. The CSI feedback method according to claim 2, wherein the
quantity M of the antenna groups of the antenna array varies along
with the rank r of the precoding matrix W.
6. The CSI feedback method according to claim 1, further
comprising: acquiring a channel measurement result, and determining
codebook parameters of a codebook in accordance with the channel
measurement result, a set of the first precoding matrices W.sub.1
forming a first codebook, a set of the second precoding matrices
W.sub.2 forming a second codebook, a set of the precoding matrices
for data transmission acquired after the calculation forming the
codebook; and feeding back the codebook parameters to the access
device, or determining one or more parameters corresponding to the
codebook parameters in accordance with a predetermined mapping
relationship between the codebook parameters and the one or more
parameters and feeding back the one or more parameters to the
access device.
7. The CSI feedback method according to claim 1, further
comprising: receiving one or more parameters from the access
device; and determining codebook parameters corresponding to the
one or more parameters in accordance with a predetermined mapping
relationship between the codebook parameters and the one or more
parameters, or receiving all the codebook parameters from the
access device.
8. The CSI feedback method according to claim 6, wherein the
codebook parameters comprise one or more of the quantities:
quantity M of the antenna groups of the antenna array, a grouping
mode of the antenna groups of the antenna array, the quantity
N.sub.i of the antenna elements in each antenna group, the quantity
of rows and columns of the antenna elements in each antenna group,
a sampling rate of the beams in the precoding matrix, and a
constitution mode of the beam groups for each antenna group in
W.sub.1.
9. A Channel State Information (CS) feedback method, comprising:
receiving a first precoding matrix index value and a second
precoding matrix index value from a User Equipment (UE), the first
precoding matrix index value corresponding to a first precoding
matrix W.sub.1 in a set of first precoding matrices determined by
the UE, the first precoding matrix W.sub.1 consisting of beam
groups for a plurality of antenna groups of an antenna array, the
second precoding matrix index value corresponding to a second
precoding matrix W.sub.2 in a set of second precoding matrices
determined by the UE, the second precoding matrix W.sub.2
comprising a beam selection section for selecting beams from the
beam groups for each antenna group and a phase adjustment section
for adjusting phases among the plurality of antenna groups; and
selecting corresponding precoding matrices from the set of first
precoding matrices and the set of second precoding matrices in
accordance with the first precoding matrix index value and the
second precoding matrix index value, and implementing calculation
on the selected precoding matrices to obtain a precoding matrix W
for data transmission.
10. The CSI feedback method according to claim 9, further
comprising: constructing a set of the first precoding matrices
W.sub.1 to obtain a first codebook and constructing a set of the
second precoding matrices W.sub.2 to obtain a second codebook; and
acquiring a codebook in accordance with the matrices in the first
codebook and the second codebook through calculation.
11. The CSI feedback method according to claim 10, further
comprising notifying the UE of codebook parameters of the codebook,
wherein the codebook parameters comprise one or more of the
quantities: quantity M of the antenna groups of the antenna array,
a grouping mode of the antenna groups of the antenna array, the
quantity N.sub.i of antenna elements in each antenna group, the
quantity of rows and columns of the antenna elements in each
antenna group, a sampling rate of the beams in the precoding
matrix, and a constitution mode of the beam groups for each antenna
group in W.sub.1.
12-21. (canceled)
22. A Channel State Information (CSI) feedback device, comprising:
a processor; a memory connected to the processor via a bus
interface and configured to store therein programs and data for the
operation of the processor; and a transceiver connected to the
processor and the memory via the bus interface, and configured to
receive and transmit data, wherein the processor is configured to
call and execute the programs and data stored in the memory, so as
to: determine a first precoding matrix W.sub.1 in a set of first
precoding matrices, and determine a first precoding matrix index
value corresponding to the first precoding matrix W.sub.1, the
first precoding matrix W.sub.1 consisting of beam groups for a
plurality of antenna groups of an antenna array; determine a second
precoding matrix W.sub.2 in a set of second precoding matrices, and
determine a second precoding matrix index value corresponding to
the second precoding matrix W.sub.2, the second precoding matrix
W.sub.2 comprising a beam selection section for selecting beams
from the beam groups for each antenna group and a phase adjustment
section for adjusting phases among the plurality of antenna groups;
and feed back the first precoding matrix index value and the second
precoding matrix index value to an access device, so that the
access device selects corresponding precoding matrices from the set
of first precoding matrices and the set of second precoding matrix
in accordance with the first precoding matrix index value and the
second precoding matrix index value, and implements calculation on
the selected precoding matrices to obtain a precoding matrix W for
data transmission.
23. A Channel State Information (CSI) feedback device, comprising:
a processor; a memory connected to the processor via a bus
interface and configured to store therein programs and data for the
operation of the processor; and a transceiver connected to the
processor and the memory via the bus interface, and configured to
receive and transmit data, wherein the processor is configured to
call and execute the programs and data stored in the memory, so as
to: receive a first precoding matrix index value and a second
precoding matrix index value from a User Equipment (UE), the first
precoding matrix index value corresponding to a first precoding
matrix W.sub.1 in a set of first precoding matrices determined by
the UE, the first precoding matrix W.sub.1 consisting of beam
groups for a plurality of antenna groups of an antenna array, the
second precoding matrix index value corresponding to a second
precoding matrix W.sub.2 in a set of second precoding matrices
determined by the UE, the second precoding matrix W.sub.2
comprising a beam selection section for selecting beams from the
beam groups for each antenna group and a phase adjustment section
for adjusting phases among the plurality of antenna groups; and
select corresponding precoding matrices from the set of first
precoding matrices and the set of second precoding matrices in
accordance with the first precoding matrix index value and the
second precoding matrix index value, and implement calculation on
the selected precoding matrices to obtain a precoding matrix W for
data transmission.
24. The CSI feedback device according to claim 22, wherein the
precoding matrix W is calculated using the following equation: W =
W 1 .times. W 2 = [ X 0 0 X 1 X 3 0 X M - 1 ] W 2 ; ##EQU00090##
the first precoding matrix W.sub.1 is a block diagonal matrix
comprising elements of 0 other than diagonal blocks where a beam
group matrix X.sub.i is located, i=0, 1, . . . , M-1, M represents
the quantity of the antenna groups for the antenna array, X.sub.i
represents a beam group for an i.sup.th antenna group and it is a
N.sub.i.times.L.sub.i matrix with each column representing one beam
direction, N.sub.i represents the quantity of antenna elements in
the i.sup.th antenna group, N represents a total quantity of
antenna elements for the antenna array and N = i = 0 M - 1 .times.
N i , ##EQU00091## and L.sub.i represents the quantity of beams in
the beam group for the i.sup.th antenna group; the second precoding
matrix W.sub.2 is used for beam selection and phase adjustment, and
it is capable of selecting several beams from each beam group and
adjusting phases among the selected beams for each antenna group,
to acquire a final precoding matrix; and the second precoding
matrix W.sub.2 is a ( i = 0 M - 1 .times. L i ) .times. r
##EQU00092## matrix, where r represents a rank of the precoding
matrix W.
25. The CSI feedback device according to claim 24, wherein W 2 ' =
[ E 0 0 E 1 E 2 0 E M - 1 ] W p , where .times. [ E 1 0 E 1 E 2 0 E
M - 1 ] ##EQU00093## represents the beam selection section and is
to select r.sub.1 beams from each antenna group, a beam selection
vector for the i.sup.th antenna group is defined as E.sub.i=.left
brkt-bot.e.sub.i,0 e.sub.i,1 . . . e.sub.i,r.sub.1.sub.-1.right
brkt-bot., where e.sub.i,j represents a column selection vector
having a length of L.sub.i and it is to select one beam from the
beam group matrix X.sub.i or merge the beams in X.sub.i into one
beam, j=0, 1, . . . r.sub.1-1, W.sub.p represents the phase
adjustment section, and r columns of W.sub.2' are selected as the
second precoding matrix W.sub.2.
26. The CSI feedback device according to claim 25, wherein when the
r.sub.1 beams are orthogonal to each other, W.sub.p=UI.sub.r.sub.1,
where U represents an orthogonal matrix comprising r.sub.2 columns,
each column comprises M phase adjustment factors, r.sub.2 is a
minimum integer greater than or equal to r r 1 , ##EQU00094##
I.sub.r.sub.1 represents a r.sub.1.times.r.sub.1 unit matrix; and
when the r.sub.1 beams are not orthogonal to each other, every
r.sub.1 groups of M antenna groups form M r 1 ##EQU00095## superior
groups, and W.sub.p=UV, where U represents an orthogonal matrix for
adjusting phases among the superior groups and comprising r.sub.2
columns, each column comprises M r 1 ##EQU00096## phase adjustment
factors, r.sub.2 is a minimum integer greater than or equal to r r
1 , ##EQU00097## V is for adjusting phases among the antenna groups
in one superior group and V=.left
brkt-bot.V'(:,1)e.sub.r.sub.1.sup.(1) V'(:,2)e.sub.r.sub.1.sup.(2)
. . . V'(:,r.sub.1)e.sub.r.sub.1.sup.(r.sup.1.sup.).right
brkt-bot., V' represents an orthogonal matrix comprising r.sub.1
columns, each column comprises r.sub.1 phase adjustment factors,
V'(:,i) represents an i.sup.th column of V', and
e.sub.r.sub.1.sup.(i) represents an i.sup.th column of a unit
matrix I.sub.r.sub.1.sub..times.r.sub.1.
27. The CSI feedback device according to claim 24, wherein the
quantity M of the antenna groups of the antenna array varies along
with the rank r of the precoding matrix W.
28. The CSI feedback device according to claim 22, further
comprising a first codebook parameter processing module configured
to acquire a channel measurement result, determine codebook
parameters of a codebook in accordance with the channel measurement
result, and feed back the codebook parameters to the access device,
or determine one or more parameters corresponding to the codebook
parameters in accordance with a predetermined mapping relationship
between the codebook parameters and the one or more parameters and
feed back the one or more parameters to the access device, wherein
a set of the first precoding matrices W.sub.1 form a first
codebook, a set of the second precoding matrices W.sub.2 form a
second codebook, and a set of the precoding matrices for data
transmission acquired after the calculation form the codebook.
29. The CSI feedback device according to claim 22, further
comprising a second codebook parameter processing module configured
to receive one or more parameters from the access device, and
determine codebook parameters corresponding to the one or more
parameters in accordance with a predetermined mapping relationship
between the codebook parameters and the one or more parameters or
receive all the codebook parameters from the access device.
30. The CSI feedback device according to claim 28, wherein the
codebook parameters comprise one or more of the quantities:
quantity M of the antenna groups of the antenna array, a grouping
mode of the antenna groups of the antenna array, the quantity
N.sub.i of the antenna elements in each antenna group, the quantity
of rows and columns of the antenna elements in each antenna group,
a sampling rate of the beams in the precoding matrix, and a
constitution mode of the beam groups for each antenna group in
W.sub.1.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] The present application claims a priority of the Chinese
patent application No. 201610140712.2 filed on Mar. 11, 2016, which
is incorporated herein by reference in its entirety.
TECHNICAL FIELD
[0002] The present disclosure relates to the field of communication
technology, in particular to a Channel State Information (CSI)
feedback method and a CSI feedback device.
BACKGROUND
[0003] In a Long Term Evolution (LTE) Release-10 (Rel-10) system,
two levels of codebooks are introduced with respect to 8 antenna
ports at a base station end. In an LTE Rel-12 system, two levels of
codebooks are also introduced with respect to 4 antenna ports at
the base station end. In an LTE Rel-13 system, a Full Dimension
Multi-Input-Multi-Output (FD MIMO) antenna array (as shown in FIGS.
1 and 2) is adopted at the base station end with 16 antenna ports
arranged in a one-dimension or two-dimensional manner.
[0004] Based on the configuration of the antenna array, for a
codebook design, the original two levels of codebooks with respect
to the 4 antenna ports and the 8 antenna ports are extended to
two-dimensional, two levels of codebooks. This codebook structure
may be directly extended so as to be used for more antenna
ports.
[0005] In a multi-antenna system on the basis of a linear precoding
technique, a receiver selects an appropriate precoding matrix from
a predefined set of precoding matrices in accordance with channel
information and transmits an index of the selected precoding matrix
in the set to a transmitter, and this set is just called as
codebook. The transmitter determines the corresponding precoding
matrix in accordance with the received index, and performs
appropriate preprocessing on a signal to be transmitted therefrom,
so as to improve the validity and reliability of the information
transmission. During the above procedure, the codebook is an
essential element. The codebook needs to be designed in such a
manner as to match a channel distribution characteristic and
minimize the performance loss caused by codebook quantization as
possible.
[0006] However, there is at least the following defect in the
related art. The above-mentioned codebook structure is designed by
taking the antenna array as a whole, and a difference between
channels for different sections of the antenna array is not taken
into consideration. During the feedback of CSI on the basis of the
codebook, certain performance loss may occur due to the inaccurate
channel quantization.
SUMMARY
[0007] An object of the present disclosure is to provide a CSI
feedback method and a CSI feedback device, so as to divide an
antenna array into more groups on the basis of a division mode of
the antenna array, and transmit information in a two-level feedback
mode, thereby to quantize the channel in a more accurate manner and
improve the system performance.
[0008] In one aspect, the present disclosure provides in some
embodiments a CSI feedback method, including: determining a first
precoding matrix W.sub.1 in a set of first precoding matrices, and
determining a first precoding matrix index value corresponding to
the first precoding matrix W.sub.1, the first precoding matrix
W.sub.1 consisting of beam groups for a plurality of antenna groups
of an antenna array; determining a second precoding matrix W.sub.2
in a set of second precoding matrices, and determining a second
precoding matrix index value corresponding to the second precoding
matrix W.sub.2, the second precoding matrix W.sub.2 including a
beam selection section for selecting beams from the beam groups for
each antenna group and a phase adjustment section for adjusting
phases among the plurality of antenna groups; and feeding back the
first precoding matrix index value and the second precoding matrix
index value to an access device, selecting, by the access device,
corresponding precoding matrices from the set of first precoding
matrices and the set of second precoding matrix in accordance with
the first precoding matrix index value and the second precoding
matrix index value, and implementing calculation on the selected
precoding matrices to obtain a precoding matrix W for data
transmission.
[0009] In a possible embodiment of the present disclosure, the
precoding matrix W is calculated using the following equation:
W = W 1 .times. W 2 = [ X 0 0 X 1 X 2 0 X M - 1 ] W 2 .
##EQU00001##
The first precoding matrix W.sub.1 is a block diagonal matrix
including elements of 0 other than diagonal blocks where a beam
group matrix X.sub.i is located, i=0, 1, . . . , M-1, M represents
the quantity of the antenna groups for the antenna array, X.sub.i
represents a beam group for an i.sup.th antenna group and it is a
N.sub.i.times.L.sub.i matrix with each column representing one beam
direction, N.sub.i represents the quantity of antenna elements in
the i.sup.th antenna group, N represents a total quantity of
antenna elements for the antenna array and
N = i = 0 M - 1 .times. N l , ##EQU00002##
and L.sub.i represents the quantity of beams in the beam group for
the i.sup.th antenna group. The second precoding matrix W.sub.2 is
used for beam selection and phase adjustment, and it is capable of
selecting several beams from each beam group and adjusting a phase
among the selected beams for each antenna group, to acquire a final
precoding matrix. The second precoding matrix W.sub.2 is a
( i = 0 M - .1 .times. L i ) .times. r ##EQU00003##
matrix, where r represents a rank of the precoding matrix W.
[0010] In a possible embodiment of the present disclosure,
W 2 ' = [ E 0 0 E 1 E 2 0 E M - 1 ] W p , .times. where .times. [ E
0 0 E 1 E 2 0 E M - 1 ] ##EQU00004##
represents the beam selection section and is to select r.sub.1
beams from each antenna group, a beam selection vector for the
i.sup.th antenna group is defined as E.sub.i=.left
brkt-bot.e.sub.i,0 e.sub.i,1 . . . e.sub.i,r.sub.1.sub.-1.right
brkt-bot., where e.sub.i,j represents a column selection vector
having a length of L.sub.i and it is to select one beam from the
beam group matrix X.sub.i or merge the beams in X.sub.i into one
beam, j=0, 1, . . . r.sub.1-1, W.sub.p represents the phase
adjustment section, and r columns of W.sub.2' are selected as the
second precoding matrix W.sub.2.
[0011] In a possible embodiment of the present disclosure, when the
r.sub.1 beams are orthogonal to each other, W.sub.p=UI.sub.r.sub.1,
where U represents an orthogonal matrix including r.sub.2 columns,
each column includes M phase adjustment factors, r.sub.2 is a
minimum integer greater than or equal to
r r 1 , ##EQU00005##
I.sub.r.sub.1 represents a r.sub.1.times.r.sub.1 unit matrix. When
the r.sub.1 beams are not orthogonal to each other, every r.sub.1
groups of M antenna groups form
M r 1 ##EQU00006##
superior groups, and W.sub.p=UV, where U represents an orthogonal
matrix for adjusting phases among the superior groups and including
r.sub.2 columns, each column includes
M r 1 ##EQU00007##
phase adjustment factors, r.sub.2 is a minimum integer greater than
or equal to
r r 1 , ##EQU00008##
V is used for adjusting a phase among the antenna groups in one
superior group and V=.left brkt-bot.V'(:,1)e.sub.r.sub.1.sup.(1)
V'(:,2)e.sub.r.sub.1.sup.(2) . . .
V'(:,r.sub.1)e.sub.r.sub.1.sup.(r.sup.1.sup.).right brkt-bot., V'
represents an orthogonal matrix including r.sub.1 columns, each
column includes r.sub.1 phase adjustment factors, V'(:,i)
represents an i.sup.th column of V', and e.sub.r.sub.1.sup.(i)
represents an i.sup.th column of a unit matrix
I.sub.r.sub.1.sub..times.r.sub.1.
[0012] In a possible embodiment of the present disclosure, the
quantity M of the antenna groups of the antenna array varies along
with the rank r of the precoding matrix W.
[0013] In a possible embodiment of the present disclosure, the CSI
feedback method further includes: acquiring a channel measurement
result, and determining codebook parameters of a codebook in
accordance with the channel measurement result, a set of the first
precoding matrices W.sub.1 forming a first codebook, a set of the
second precoding matrices W.sub.2 forming a second codebook, a set
of the precoding matrices for data transmission acquired after the
calculation forming the codebook; and feeding back the codebook
parameters to the access device, or determining one or more
parameters corresponding to the codebook parameters in accordance
with a predetermined mapping relationship between the codebook
parameters and the one or more parameters and feeding back the one
or more parameters to the access device.
[0014] In a possible embodiment of the present disclosure, the CS
feedback method further includes: receiving one or more parameters
from the access device; and determining codebook parameters
corresponding to the one or more parameters in accordance with a
predetermined mapping relationship between the codebook parameters
and the one or more parameters, or receiving all the codebook
parameters from the access device.
[0015] In a possible embodiment of the present disclosure, the
codebook parameters include one or more of the quantities: quantity
M of the antenna groups of the antenna array, a grouping mode of
the antenna groups of the antenna array, the quantity N.sub.i of
the antenna elements in each antenna group, the quantity of rows
and columns of the antenna elements in each antenna group, a
sampling rate of the beams in the precoding matrix, and a
constitution mode of the beam groups for each antenna group in
W.sub.1.
[0016] In another aspect, the present disclosure provides in some
embodiments a CSI feedback method, including: receiving a first
precoding matrix index value and a second precoding matrix index
value from a User Equipment (UE), the first precoding matrix index
value corresponding to a first precoding matrix W.sub.1 in a set of
first precoding matrices determined by the UE, the first precoding
matrix W.sub.1 consisting of beam groups for a plurality of antenna
groups of an antenna array, the second precoding matrix index value
corresponding to a second precoding matrix W.sub.2 in a set of
second precoding matrices determined by the UE, the second
precoding matrix W.sub.2 including a beam selection section for
selecting beams from the beam groups for each antenna group and a
phase adjustment section for adjusting phases among the plurality
of antenna groups; and selecting corresponding precoding matrices
from the set of first precoding matrices and the set of second
precoding matrices in accordance with the first precoding matrix
index value and the second precoding matrix index value, and
implementing calculation on the selected precoding matrices to
obtain a precoding matrix W for data transmission.
[0017] In a possible embodiment of the present disclosure, the CSI
feedback method further includes: constructing a set of the first
precoding matrices W.sub.1 to obtain a first codebook and
constructing a set of the second precoding matrices W.sub.2 to
obtain a second codebook; and acquiring a codebook in accordance
with the matrices in the first codebook and the second codebook
through calculation.
[0018] In a possible embodiment of the present disclosure, the CSI
feedback method further includes notifying the UE of codebook
parameters of the codebook. The codebook parameters include one or
more of the quantities: quantity M of the antenna groups of the
antenna array, a grouping mode of the antenna groups of the antenna
array, the quantity N.sub.i of antenna elements in each antenna
group, the quantity of rows and columns of the antenna elements in
each antenna group, a sampling rate of the beams in the precoding
matrix, and a constitution mode of the beam groups for each antenna
group in W.sub.1.
[0019] In yet another aspect, the present disclosure provides in
some embodiments a CSI feedback device, including: a first
determination module configured to determine a first precoding
matrix W.sub.1 in a set of first precoding matrices, and determine
a first precoding matrix index value corresponding to the first
precoding matrix W.sub.1, the first precoding matrix W.sub.1
consisting of beam groups for a plurality of antenna groups of an
antenna array; a second predetermined module configured to
determine a second precoding matrix W.sub.2 in a set of second
precoding matrices, and determine a second precoding matrix index
value corresponding to the second precoding matrix W.sub.2, the
second precoding matrix W.sub.2 including a beam selection section
for selecting beams from the beam groups for each antenna group and
a phase adjustment section for adjusting phases among the plurality
of antenna groups; and a first feedback module configured to feed
back the first precoding matrix index value and the second
precoding matrix index value to an access device, so as to enable
the access device to select corresponding precoding matrices from
the set of first precoding matrices and the set of second precoding
matrix in accordance with the first precoding matrix index value
and the second precoding matrix index value, and implement
calculation on the selected precoding matrices to obtain a
precoding matrix W for data transmission.
[0020] In a possible embodiment of the present disclosure, the
precoding matrix W is calculated using the following equation:
W = W 1 .times. W 2 = [ X 0 0 X 1 X 2 0 X M - 1 ] W 2 .
##EQU00009##
The first precoding matrix W.sub.1 is a block diagonal matrix
including elements of 0 other than diagonal blocks where a beam
group matrix X.sub.i is located, i=0, 1, . . . , M-1, M represents
the quantity of the antenna groups for the antenna array, X.sub.i
represents a beam group for an i.sup.th antenna group and it is a
N.sub.i.times.L.sub.i matrix with each column representing one beam
direction, N.sub.i represents the quantity of antenna elements in
the i.sup.th antenna group, N represents a total quantity of
antenna elements for the antenna array and
N = i = 0 M - 1 .times. N i , ##EQU00010##
and L.sub.i represents the quantity of beams in the beam group for
the i.sup.th antenna group. The second precoding matrix W.sub.2 is
used for beam selection and phase adjustment, and it is capable of
selecting several beams from each beam group and adjusting a phase
among the selected beams for each antenna group, to acquire a final
precoding matrix. The second precoding matrix W.sub.2 is a
( i = 0 M - 1 .times. L i ) .times. r ##EQU00011##
matrix, where r represents a rank of the precoding matrix W.
[0021] In a possible embodiment of the present disclosure,
W 2 ' = [ E 0 0 E 1 E 2 0 E M - 1 ] W p , where .times. [ E 0 0 E 1
E 2 0 E M - 1 ] ##EQU00012##
represents the beam selection section and is to select r.sub.1
beams from each antenna group, a beam selection vector for the
i.sup.th antenna group is defined as E.sub.i=.left
brkt-bot.e.sub.i,0 e.sub.i,1 . . . e.sub.i,r.sub.1.sub.-1.right
brkt-bot., where e.sub.i,j represents a column selection vector
having a length of L.sub.i and it is to select one beam from the
beam group matrix X.sub.i or merge the beams in X.sub.i into one
beam, j=0, 1, . . . r.sub.1-1, W.sub.p represents the phase
adjustment section, and r columns of W.sub.2' are selected as the
second precoding matrix W.sub.2.
[0022] In a possible embodiment of the present disclosure, when the
r.sub.1 beams are orthogonal to each other, W.sub.p=UI.sub.r.sub.1,
where U represents an orthogonal matrix including r.sub.2 columns,
each column includes M phase adjustment factors, r.sub.2 is a
minimum integer greater than or equal to
r r 1 , ##EQU00013##
I.sub.r.sub.1 represents a r.sub.1.times.r.sub.1 unit matrix. When
the r.sub.1 beams are not orthogonal to each other, every r.sub.1
groups of M antenna groups form
M r 1 ##EQU00014##
superior groups, and W.sub.p=UV, where U represents an orthogonal
matrix for adjusting a phase among the superior groups and
including r.sub.2 columns, each column includes
M r 1 ##EQU00015##
phase adjustment factors, r.sub.2 is a minimum integer greater than
or equal to
r r 1 , ##EQU00016##
V is for adjusting phases among the antenna groups in one superior
group and V=.left brkt-bot.V'(:,1)e.sub.r.sub.1.sup.(1)
V'(:,2)e.sub.r.sub.1.sup.(2) . . .
V'(:,r.sub.1)e.sub.r.sub.1.sup.(r.sup.1.sup.).right brkt-bot., V'
represents an orthogonal matrix including r.sub.1 columns, each
column includes r.sub.1 phase adjustment factors, V'(:,i)
represents an i.sup.th column of V', and e.sub.r.sub.1.sup.(i)
represents an i.sup.th column of a unit matrix
I.sub.r.sub.1.sub..times.r.sub.1.
[0023] Ina possible embodiment of the present disclosure, the
quantity M of the antenna groups of the antenna array varies along
with the rank r of the precoding matrix W.
[0024] In a possible embodiment of the present disclosure, the CSI
feedback device further includes a first codebook parameter
processing module configured to acquire a channel measurement
result, determine codebook parameters of a codebook in accordance
with the channel measurement result, and feedback the codebook
parameters to the access device, or determine one or more
parameters corresponding to the codebook parameters in accordance
with a predetermined mapping relationship between the codebook
parameters and the one or more parameters and feedback the one or
more parameters to the access device. A set of the first precoding
matrices W.sub.1 form a first codebook, a set of the second
precoding matrices W.sub.2 form a second codebook, and a set of the
precoding matrices for data transmission acquired after the
calculation form the codebook.
[0025] In a possible embodiment of the present disclosure, the CSI
feedback device further includes a second codebook parameter
processing module configured to receive one or more parameters from
the access device, and determine codebook parameters corresponding
to the one or more parameters in accordance with a predetermined
mapping relationship between the codebook parameters and the one or
more parameters or receive all the codebook parameters from the
access device.
[0026] In a possible embodiment of the present disclosure, the
codebook parameters include one or more of the quantities: quantity
M of the antenna groups of the antenna array, a grouping mode of
the antenna groups of the antenna array, the quantity N.sub.i of
the antenna elements in each antenna group, the quantity of rows
and columns of the antenna elements in each antenna group, a
sampling rate of the beams in the precoding matrix, and a
constitution mode of the beam groups for each antenna group in
W.sub.1.
[0027] In still yet another aspect, the present disclosure provides
in some embodiments a CSI feedback device, including: a second
feedback module configured to receive a first precoding matrix
index value and a second precoding matrix index value from a UE,
the first precoding matrix index value corresponding to a first
precoding matrix W.sub.1 in a set of first precoding matrices
determined by the UE, the first precoding matrix W.sub.1 consisting
of beam groups for a plurality of antenna groups of an antenna
array, the second precoding matrix index value corresponding to a
second precoding matrix W.sub.2 in a set of second precoding
matrices determined by the UE, the second precoding matrix W.sub.2
including a beam selection section for selecting beams from the
beam groups for each antenna group and a phase adjustment section
for adjusting a phase among the plurality of antenna groups; and a
precoding matrix selection module configured to select
corresponding precoding matrices from the set of first precoding
matrices and the set of second precoding matrices in accordance
with the first precoding matrix index value and the second
precoding matrix index value, and implement calculation on the
selected precoding matrices to obtain a precoding matrix W for
data.
[0028] In a possible embodiment of the present disclosure, the CSI
feedback device further includes a third feedback module configured
to notify the UE of codebook parameters of a codebook. The codebook
parameters include one or more of the quantities: quantity M of the
antenna groups of the antenna array, a grouping mode of the antenna
groups of the antenna array, the quantity N.sub.i of antenna
elements in each antenna group, the quantity of rows and columns of
the antenna elements in each antenna group, a sampling rate of the
beams in the precoding matrix, and a constitution mode of the beam
groups for each antenna group in W.sub.1. A set of the first
precoding matrices W.sub.1 form a first codebook, a set of the
second precoding matrices W.sub.2 form a second codebook, and a set
of the precoding matrices for data transmission acquired after the
calculation form the codebook.
[0029] According to the embodiments of the present disclosure, the
antenna array is divided into a plurality of antenna groups.
Through two levels of codebooks, it is able to achieve the beam
selection in each antenna group and the phase adjustment among the
antenna groups. The two levels of codebooks are capable of being
fed back respectively. As a result, it is able to quantize the
channel in a more accurate manner, and improve the system
performance. In addition, the quantity of the antenna groups may
vary along with the quantity of data streams.
BRIEF DESCRIPTION OF THE DRAWINGS
[0030] FIG. 1 is a schematic view showing 12 antenna ports for a
two-dimensional antenna array in an LTE Rel-13 system;
[0031] FIG. 2 is a schematic view showing 16 antenna ports for the
two-dimensional antenna array in the LTE Rel-13 system;
[0032] FIG. 3 is a flow chart of a CSI feedback method according to
a first embodiment of the present disclosure;
[0033] FIG. 4A is a schematic view showing antenna groups of a
single-polarized antenna array according to the first embodiment of
the present disclosure;
[0034] FIG. 4B is a schematic view showing antenna groups of a
dual-polarized antenna array according to the first embodiment of
the present disclosure;
[0035] FIG. 5 is a flow chart of a CSI feedback method according to
a second embodiment of the present disclosure;
[0036] FIG. 6 is a block diagram of a CSI feedback device according
to a third embodiment of the present disclosure; and
[0037] FIG. 7 is a block diagram of a CSI feedback device according
to a fourth embodiment of the present disclosure.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0038] In the embodiments of the present disclosure, a UE may be a
mobile phone or any other device capable of transmitting or
receiving a radio signal, including a terminal, a Personal Digital
Assistant (PDA), a wireless modem, a wireless communication device,
a handheld device, a laptop computer, a cordless telephone, a
Wireless Local Loop (WLL) station, Customer Premise Equipment (CPE)
or a mobile intelligent hot spot capable of converting a mobile
signal into a Wireless Fidelity (WiFi) signal, intelligent
household electrical appliance, or any other device capable of
spontaneously communicating with a mobile communication network. An
access device may be a base station. It should be appreciated that,
the form of the base station will not be particularly defined
herein, and it may be a macro base station, a pico base station,
NodeB (a 3.sup.rd-Generation (3G) mobile station), an evolved NodeB
(eNB), a femto eNB (or Home eNB (HeNB)), a relay, an access point,
a Remote Radio Unit (RRU), or a Remote Radio Head (RRH).
First Embodiment
[0039] The present disclosure provides in this embodiment a CSI
feedback method implemented by a UE. As shown in FIG. 3, the CSI
feedback method includes the following steps.
[0040] Step S301: determining a first precoding matrix W.sub.1 in a
set of first precoding matrices, and determining a first precoding
matrix index value (i.sub.1,1,i.sub.1,2) corresponding to the first
precoding matrix W.sub.1. The first precoding matrix W.sub.1
consists of beam groups for a plurality of antenna groups of an
antenna array.
[0041] Step 302: determining a second precoding matrix W.sub.2 in a
set of second precoding matrices, and determining a second
precoding matrix index value (i.sub.1) corresponding to the second
precoding matrix W.sub.2. The second precoding matrix W.sub.2
includes a beam selection section for selecting beams from the beam
groups for each antenna group and a phase adjustment section for
adjusting phases among the plurality of antenna groups.
[0042] In a possible embodiment of the present disclosure,
different granularities of feedback frequency domains and feedback
periods may be adopted by the first precoding matrix W.sub.1 and
the second precoding matrix W.sub.2. For example, the first
precoding matrix W.sub.1 is for describing a long-term
characteristic of a channel, so it may be fed back using a wide
band in a long-time manner. The second precoding matrix W.sub.2 is
for describing a short-term characteristic of the channel, so it
may be fed back using a subband in a short-term manner.
[0043] Step S303: feeding back the first precoding matrix index
value (i.sub.1,1,i.sub.1,2) and the second precoding matrix index
value (i.sub.2) to an access device, so that the access device
selects corresponding precoding matrices from the set of first
precoding matrices and the set of second precoding matrix in
accordance with the first precoding matrix index value
(i.sub.1,1,i.sub.1,2) and the second precoding matrix index value
(i.sub.2), and implements calculation on the selected precoding
matrices to obtain a precoding matrix W for data transmission.
[0044] It should be appreciated that, a set of the precoding
matrices in the above form is called as a codebook. In a possible
embodiment of the present disclosure, the CSI feedback method
further includes a step of configuring codebook parameters of the
codebook. To be specific, this step includes: acquiring a channel
measurement result, determining the codebook parameters of the
codebook in accordance with the channel measurement result, and
transmitting the codebook parameters to the access device; or
acquiring the channel measurement result, determining the codebook
parameters of the codebook in accordance with the channel
measurement result, determining one or more parameters
corresponding to the codebook parameters in accordance with a
predetermined mapping relationship between the codebook parameters
and the one or more parameters, and transmitting the one or more
parameters to the access device.
[0045] The codebook parameters include one or more of the
quantities: quantity M of the antenna groups, a grouping mode of
the antenna groups, the quantity N.sub.i of antenna elements in
each antenna group, the quantity of rows and columns of the antenna
elements in each antenna group, a sampling rate of the beams in a
precoding matrix, and a constitution mode of the beam groups (a
beam group configuration mode) for each antenna group in W.sub.1.
The mapping relationship may be a mapping relation between the
codebook parameters and the quantity of antenna ports, or between
the codebook parameters and the quantity of the antenna ports as
well as the quantity of data streams.
[0046] The present disclosure will be described hereinafter in
conjunction with specific scenarios.
[0047] The antenna elements of the antenna array may be grouped.
FIG. 4A shows the antenna groups of a single-polarized antenna
array, where the antenna elements within each dotted box form a
group. FIG. 4B shows the antenna groups of a dual-polarized antenna
array, where the antenna elements within each dotted box form two
groups and the antenna elements in each polarization direction form
one group. The quantity of the antenna groups of the antenna array
may be represented by M.
[0048] The precoding matrix W adopted by the access device has the
following form:
W = W 1 .times. W 2 = [ X 0 0 X 1 X 2 0 X M - 1 ] W 2 .
##EQU00017##
The first precoding matrix W.sub.1 is a block diagonal matrix
including elements of 0 other than diagonal blocks where a beam
group matrix X.sub.i is located, i=0, 1, . . . , M-1, M represents
the quantity of the antenna groups for the antenna array, X.sub.i
represents a beam group for an i.sup.th antenna group and it is a
N.sub.i.times.L.sub.i matrix, N.sub.i represents the quantity of
antenna elements in the i.sup.th antenna group, N represents a
total quantity of antenna elements for the antenna array and
N = i = 0 M - 1 .times. N i , ##EQU00018##
and L.sub.i represents the quantity of beams in the beam group for
the i.sup.th antenna group. In a possible embodiment of the present
disclosure, the beam group matrix X.sub.i may consist of a group of
Discrete Fourier Transform (DFT) vectors, and each DFT vector
represents a beam direction.
[0049] The second precoding matrix W.sub.2 is configured for beam
selection and phase adjustment, and it is capable of selecting
several beams from each beam group and adjusting phases of the
selected beams for respective antenna groups, so as to acquire a
final precoding matrix.
[0050] In a possible embodiment of the present disclosure,
W 2 ' = [ E 0 0 E 1 E 2 0 E M - 1 ] W p , where .times. [ E 0 0 E 1
E 2 0 E M - 1 ] ##EQU00019##
represents the beam selection section and configured to select
r.sub.1 beams from each antenna group, a beam selection vector for
the i.sup.th antenna group is defined as E.sub.i=.left
brkt-bot.e.sub.i,0 e.sub.i,1 . . . e.sub.i,r.sub.1.sub.-1.right
brkt-bot., where e.sub.i,j represents a column selection vector
having a length of L.sub.i and it is configured to select one beam
from the beam group matrix X.sub.i or merge the beams in X.sub.i
into one beam, j=0, 1, . . . r.sub.1-1, W.sub.p represents the
phase adjustment section, and r columns of W.sub.2' are selected as
the second precoding matrix W.sub.2.
[0051] In a possible embodiment of the present disclosure, merely
one element in e.sub.i,j is 1, and the other elements are 0.
[0052] a) When the r.sub.1 beams are orthogonal to each other, the
phase adjustment among the antenna groups may be defined as U=.left
brkt-bot.u.sub.0 u.sub.1 . . . u.sub.r.sub.2.sub.-1.right
brkt-bot., where r.sub.2 is a minimum integer greater than or equal
to
r r 1 , ##EQU00020##
u.sub.i represents a phase adjustment factor among the antenna
groups for adjusting a phase relationship among the antenna groups
and it is a column vector having a length of M, and column vectors
in U are orthogonal to each other so as to ensure that there is no
interference among the data streams. In a possible embodiment of
the present disclosure, U is generated through Householder
transform. Hence, W.sub.2' may be expressed as:
W 2 ' = [ E 0 0 E 1 E 2 0 E M - 1 ] ( U I r 1 ) , ##EQU00021##
where I.sub.r.sub.1 is a r.sub.1.times.r.sub.1 unit matrix. r
columns, e.g., previous r columns, may be selected from W.sub.2' as
W.sub.2. It should be appreciated that, when calculating the final
precoding matrix, a result of W.sub.1W.sub.2 may be normalized.
[0053] When r.sub.1=1,
W 2 ' = [ E 0 0 E 1 E 2 0 E M - 1 ] U . ##EQU00022##
Further, when E.sub.0=E.sub.1= . . . =E.sub.M-1,
W.sub.2=UE.sub.0.
[0054] b) When the r.sub.1 beams are not orthogonal to each other,
a same beam or different beams may be selected.
[0055] To be specific, when the r.sub.1 beams are not orthogonal to
each other, these r.sub.1 beams may be the same or different from
each other, on the premise that M is capable of being divided
exactly by r.sub.1. At this time, every r.sub.1 antenna groups of
the M antenna groups may form a large group, i.e., a superior
group, so there are totally
M r 1 ##EQU00023##
superior groups. In a possible embodiment of the present
disclosure, the same r.sub.1 beams may be selected for the antenna
groups in each superior group. W.sub.p=UV, where U is an orthogonal
matrix for the phase adjustment among the superior groups and
includes r.sub.2 columns each including
M r 1 ##EQU00024##
phase adjustment factors, r.sub.2 is a minimum integer greater than
or equal to
r r 1 , ##EQU00025##
V is configured for the phase adjustment among the antenna groups
in one superior group and V=.left
brkt-bot.V'(:,1)e.sub.r.sub.1.sup.(1) V'(:,2)e.sub.r.sub.1.sup.(2)
. . . V'(:,r.sub.1)e.sub.r.sub.1.sup.(r.sup.1.sup.).right
brkt-bot., V' is an orthogonal matrix and includes r.sub.1 columns
each including r phase adjustment factors, V'(:,i) represents an
i.sup.th column of V', and e.sub.r.sub.1.sup.(i) represents an
i.sup.th column of a unit matrix
I.sub.r.sub.1.sub..times.r.sub.1.
[0056] Based on the above, the rank r of the precoding matrix
consists of two parts, i.e., the quantity r.sub.1 of the beams
selected for each antenna group, and the quantity r.sub.2 of the
data streams among the antenna groups.
[0057] When r.sub.1=1 and r.sub.2=r, merely one beam may be
selected for each antenna group, r data streams may be transmitted
parallel to each other among the antenna groups through the phase
adjustment. Usually, the distance between antennas in the antenna
group is relatively small and the correlation of the antenna array
is relatively strong, so low-rank transmission may be performed
appropriately. The distance between the antenna groups is
relatively large and the correlation of the antenna array is
relatively weak, so high-rank transmission may be performed
appropriately. This configuration is relatively typical.
[0058] When r.sub.1=r and r.sub.2=1, r beams may be selected from
each antenna group for the parallel transmission of r data streams,
and the phase adjustment among the antenna groups may be merely
performed with respect to these data streams.
[0059] When r.sub.1=2 and
r 2 = r + 1 2 , ##EQU00026##
two data streams may be transmitted parallel to each other in each
antenna group, and
r r 1 ##EQU00027##
or more data streams may be transmitted through the phase
adjustment among the antenna groups. For the other combination
modes of r.sub.1 and r.sub.2, the beam selection and the phase
adjustment may be performed in a similar way.
[0060] In a possible embodiment of the present disclosure, the
quantity M of the antenna groups of the antenna array may vary
along with the quantity of the data streams (i.e., the rank r of
the precoding matrix). For example, a relationship between the
quantity M of the antenna groups and the quantity of the data
streams needs to meet a condition M.gtoreq.r.sub.2, where r.sub.2
is a minimum integer greater than or equal to
r r i . ##EQU00028##
[0061] In a possible embodiment of the present disclosure, the
first precoding matrix W.sub.1 is configured for describing a
long-term characteristic of a channel, so it may be fed back using
a wide band in a long-time manner. The second precoding matrix
W.sub.2 is for describing a short-term characteristic of the
channel, so it may be fed back using a subband in a short-term
manner.
[0062] In a possible embodiment of the present disclosure, a set of
the first precoding matrices W.sub.1 and a set of the second
precoding matrices W.sub.2 may be are constructed respectively, and
then the code book may be calculated in accordance with the
matrices in the two sets. The set of the first precoding matrices
W.sub.1 may be called as a first-level codebook, and the set of the
second precoding matrices W.sub.2 may be called as a second-level
codebook.
Example 1
[0063] For the dual-polarized antenna array, typically r.sub.1=1,
and the antenna groups each include a same quantity of antenna
elements, i.e., N.sub.0=N.sub.1= . . . =N.sub.M-1.
[0064] In a possible embodiment of the present disclosure, the
relationship between the quantity of the antenna groups of the
antenna array and the rank r of the precoding matrix may meet a
condition
M = 2 r + 1 2 , ##EQU00029##
where r represents the rank of the precoding matrix, M represents
the quantity of the antenna groups of the antenna array, and .left
brkt-bot. .right brkt-bot. represents a round-down operation.
[0065] In another possible embodiment of the present disclosure,
the relationship between the quantity of the antenna groups of the
antenna array and the rank r of the precoding matrix is shown in
Table 1.
TABLE-US-00001 TABLE 1 relationship between the quantity of antenna
groups of antenna array and rank of precoding matrix Rank .sup.r of
precoding The quantity .sup.M of the antenna matrix groups of
antenna array 1 2 2 2 3 4 4 4 5 8 6 8 7 8 8 8
[0066] In a possible embodiment of the present disclosure, the
beams for each antenna groups may be selected from a same set of
beam vectors, and the quantity L.sub.i of the beams included in the
beam group matrices X.sub.i may be the same.
[0067] W.sub.1 may be constructed using the DFT vectors as
follows.
[0068] i) When each antenna group is a two-dimensional antenna
array, two groups of DFT vectors in two dimension may be defined as
follows:
h l = [ 1 e j .times. .times. 2 .times. .pi. .times. .times. l N k
.times. O 1 e j .times. .times. 2 .times. .pi. ( N h - 1 ) l N b
.times. O 1 ] T , l = 0 , 1 , .times. , N h .times. O 1 - 1 , and
##EQU00030## v k = [ 1 e j .times. .times. 2 .times. .pi. .times.
.times. k N v .times. O 2 e j .times. .times. 2 .times. .pi. ( N v
- 1 ) .times. k N v .times. O 2 ] T , k = 0 , 1 , .times. , N v
.times. O 2 - 1 , ##EQU00030.2##
where N.sub.h represents the quantity of the antennas in each
antenna group in a first dimension, N.sub.v represents the quantity
of antennas in each antenna group in a second dimension,
N.sub.i=N.sub.hN.sub.v, and O.sub.1 and O.sub.2 represent an
oversampling factor in the first dimension and an oversampling
factor in the second direction respectively. A Kronecker product of
the two DFT vectors in the two dimensions may form a beam
direction, so a set of N.sub.hO.sub.1N.sub.vO.sub.2 beam vectors
may be provided as follows: {k.sub.l,k=v.sub.kh.sub.l, l=0, 1, . .
. , N.sub.hO.sub.1-1, k=0, 1, . . . , N.sub.vO.sub.2-1}.
[0069] Every L.sub.i vectors in the set of beam vectors may form
one group. In a possible grouping mode, L.sub.i=4, and each beam
group may be defined using a two-dimensional index
(i.sub.1,1,i.sub.1,2), i.e.,
X.sup.(i.sup.1,1.sup.,i.sup.1,2.sup.)=[v.sub.mod(2i.sub.1,2.sub.,N.sub.v.-
sub.O.sub.2.sub.)h.sub.mod(2i.sub.1,1.sub.,N.sub.h.sub.O.sub.1.sub.),
v.sub.mod(2i.sub.1,2.sub.,N.sub.v.sub.O.sub.2.sub.)h.sub.mod(2i.sub.1,1.s-
ub.,N.sub.h.sub.O.sub.1.sub.),
v.sub.mod(2i.sub.1,2.sub.,N.sub.v.sub.O.sub.2.sub.)h.sub.mod(2i.sub.1,1.s-
ub.,N.sub.h.sub.O.sub.1.sub.),
v.sub.mod(2i.sub.1,2.sub.,N.sub.v.sub.O.sub.2.sub.)h.sub.mod(2i.sub.1,1.s-
ub.,N.sub.h.sub.O.sub.1.sub.)], where i.sub.1,1=0, 1, . . . ,
N.sub.hO.sub.1/2-1, and i.sub.1,2=0, 1, . . . , N.sub.vO.sub.2/2-1.
The entire set of beam vectors may include
(N.sub.hO.sub.1/2)(N.sub.vO.sub.2/2) beam groups. Hence, with
respect to the beam group X.sub.i for each antenna group,
X.sub.i.di-elect cons.{X.sup.(i.sup.1,1.sup.,i.sup.1,2.sup.),
i.sub.1,1=0, 1, . . . , N.sub.hO.sub.1/2-1, i.sub.1,2=0, 1, . . . ,
N.sub.vO.sub.2/2-1}.
[0070] When all the antenna groups have the same beam groups, i.e.,
X.sub.0=X.sub.1= . . . =X.sub.M-1, the first-level codebook may
include (N.sub.hO.sub.1/2)(N.sub.vO.sub.2/2) W.sub.1 codewords.
During the feedback of the CSI, the UE may feed back the beam group
index (i.sub.1,1,i.sub.1,2) as a codeword index to the access
device. Otherwise, the first-level codebook may include
((N.sub.hO.sub.1/2)(N.sub.vO.sub.2/2)).sup.M W.sub.1 codewords.
During the feedback of the CSI, the UE may feed back the beam group
index (i.sub.1,1,i.sub.1,2) for each antenna group to the access
device, or feed back indices of combinations of the M beam group
indices in the ((N.sub.hO.sub.1/2)(N.sub.vO.sub.2/2)).sup.M
codewords to the access device.
[0071] ii) When each antenna group is a one-dimension antenna
array, a group of DFT vectors may be defined as follows:
.times. v l = [ 1 e j .times. .times. 2 .times. ? N i .times. O e j
.times. .times. 2 .times. .pi. ( N i - 1 ) l N i .times. O ] T , l
= 0 , 1 , .times. , N i .times. O - 1 , .times. ? .times. indicates
text missing or illegible when filed ##EQU00031##
where N.sub.i represents the quantity of the antennas in each
antenna group, and O represents an oversampling factor. Hence, a
set of N.sub.iO beam vectors may be provided. Every L.sub.i vectors
in the set of beam vectors may form one group. In a possible
grouping mode, L.sub.i=4, and each beam group may be defined using
a one-dimension index i.sub.i, i.e.,
X.sup.(i.sup.1.sup.)=[v.sub.mod(2i.sub.1.sub.,N.sub.i.sub.O),
v.sub.mod(2i.sub.1.sub.+1,N.sub.i.sub.O),
v.sub.mod(2i.sub.1.sub.+2,N.sub.i.sub.O),
v.sub.mod(2i.sub.1.sub.+3,N.sub.i.sub.O)], where i.sub.1=0, 1, . .
. ,N.sub.iO/2-1. The entire set of beam vectors may include
N.sub.iO/2 beam groups. Hence, with respect to the beam group
X.sub.i for each antenna group, X.sub.i.di-elect
cons.{X.sup.(i.sup.1.sup.), i.sub.1=0, 1, . . . ,
N.sub.iO/2-1}.
[0072] When all the antenna groups have the same beam groups, i.e.,
X.sub.0=X.sub.1=X.sub.M-1, the first-level codebook may include
N.sub.iO/2 W.sub.1 codewords. During the feedback of the CSI, the
UE may feed back the beam group index i.sub.i as a codeword index
to the access device. Otherwise, the first-level codebook may
include (N.sub.iO/2).sup.M W.sub.1 codewords. During the feedback
of the CSI, the UE may feed back the beam group index i.sub.1 for
each antenna group to the access device, or feed back indices of
combinations of the M beam group indices in the (N.sub.iO/2).sup.M
codewords to the access device.
[0073] W.sub.2 may be constructed as follows. Because
r.sub.1=1,
W 2 = [ e 0 , 0 0 e 1 , 0 e 2 , 0 0 e M - 1 , 0 ] U ,
##EQU00032##
where e.sub.i,0 is a column selection vector having a length of 4
(because L.sub.i=4). When e.sub.i,0 is a unit vector, it may
include one element of 1 (indicating that the beams corresponding
to this column may be selected) and the other elements of 0. For
one antenna group, there are four possible column selection
sections, i.e., e.sub.i,0{[1 0 0 0].sup.T, [0 1 0 0].sup.T, [0 0 1
0].sup.T, [0 0 0 1].sup.T}. Hence, there exist 4.sup.M possible
combinations for the column selection sections of W.sub.2. For
example, , and at this time, W.sub.2=Ue.sub.0,0.
[0074] The phase adjustment matrix U among the antenna groups is an
M.times.r matrix, and it may be expressed as:
U = [ u 0 u 1 u r - 1 ] = [ u 0 , 0 u 0 , 1 u 0 , r - 1 u 1 , 0 u 1
, 1 u 1 , r - 1 u M - 1 , 0 u M - 1 , 1 u M - 1 , r - 1 ] .
##EQU00033##
Assume there are Q phase adjustment matrices, i.e., U.sub.0,
U.sub.1, U.sub.2, . . . U.sub.Q-1. As mentioned above, the elements
in each column of the matrix U are configured for adjusting the
phase among the antenna groups, and all the columns of the matrix U
need to be orthogonal to each other, so as to prevent the
interference among the data streams. In a possible embodiment of
the present disclosure, the following two construction modes may be
provided so as to meet the above requirement.
[0075] Construction Mode 1 for the Phase Adjustment Matrix Among
the Antenna Groups
[0076] As a typical method for constructing an orthogonal basis,
Householder transform may be used to construct the matrix U. A
Householder transform matrix may be defined as follows:
P.sub.n=I-2s.sub.ns.sub.n.sup.H/s.sub.n.sup.Hs.sub.n, wherein
s.sub.n represents a base vector having a length of M and each
element in s.sub.n consists of {-1, .+-.j} and {(.+-.1.+-.j)/
{square root over (2)}}, and I represents an M.times.M unit matrix.
Here, Q base vectors may be generated, i.e., n=0, 1, . . . , Q-1.
The transform matrix P.sub.n may be calculated for each base
vector, a column order of the transform matrix P.sub.n may be
adjusted, and then r columns may be selected from the transform
matrix P.sub.n, so as to acquire the phase adjustment matrix U.
Hence, there are Q combinations for the phase adjustment sections
of W.sub.2. In this regard, the second-level codebook may include
4.sup.MQ codewords, and i.sub.2 (i.sub.2=0, 1, . . . , 4.sup.MQ-1)
may be adopted as a Precoding Matrix Indicator (PMI) to indicate
and feed back the precoding matrix.
[0077] For the codebook where rank=3,
M = 2 r + 1 2 , ##EQU00034##
so M=4. Presumed that the quantity Q of the base vectors for the
Householder transform is Q=4, a set of the phase adjustment
matrices U may be expressed as
{ 1 2 .function. [ 1 1 1 1 1 - 1 1 - 1 - 1 1 - 1 1 ] 1 2 .function.
[ 1 - j - 1 j 1 j - 1 - j 1 - j 1 - j ] 1 2 .function. [ 1 - 1 1 -
1 1 1 1 1 1 - 1 - 1 1 ] 1 2 .function. [ 1 j - 1 - j 1 - j - 1 j 1
j 1 j ] } .times. ? ##EQU00035## ? .times. indicates text missing
or illegible when filed ##EQU00035.2##
[0078] Construction Mode 2 for the Phase Adjustment Matrix Among
the Antenna Groups
[0079] The phase adjustment matrix among the antenna groups may be
constructed using a Grassmannian manifold. A set of the matrices
U.sub.0, U.sub.1, U.sub.2, . . . U.sub.Q-1 may be acquired by
solving the following optimization problem
{ U 0 , U 1 , U 2 , , U Q - 1 } = arg .times. max { F 0 , F 1 ,
.times. , F Q - 1 | F q H .times. F q = I r .times. r , q = 0 , 1 ,
.times. , Q - 1 } .times. min 0 .ltoreq. i .ltoreq. j .ltoreq. Q -
1 .times. dist .function. ( F i , F j ) , ##EQU00036##
where Irepresents a r.times.r unit matrix, and dist (A, B)
represents a distance between two matrices A and B, e.g.,
dist(A,B)=.parallel.A-B.parallel.(Frobenius norm) or dist(A,
B)=.parallel.A-B.parallel..
[0080] A minimum distance among the Q matrices in the acquired set
of matrices may be of a maximum value or an approximately maximum
value in all possible sets of matrices. It should be appreciated
that, the above optimization problem may be solved through an
offline computer calculation method.
[0081] Upon the acquisition of the first-level codebook and the
second-level codebook as mentioned above, one matrix in the
first-level codebook may be multiplied by one matrix in the
second-level codebook, and then a resultant matrix may be
normalized, so as to acquire the precoding matrix in the codebook.
When each antenna group is a two-dimensional antenna array, the
normalization operation may be performed as follows:
W = 1 N h .times. N v r .function. [ X 0 0 X 1 .times. X 2 0 X M -
1 ] [ u 0 , 0 .times. e 0 , 0 u 0 , 1 .times. e 0 , 0 u 0 , r - 1
.times. e 0 , 0 u 1 , 0 .times. e 1 , 0 u 1 , 1 .times. e 1 , 0 u 1
, r - 1 .times. e 1 , 0 u 2 , 0 .times. e 2 , 0 u 2 , 1 .times. e 2
, 0 u 2 , r - 1 .times. e 2 , 0 u M - 1 , 0 .times. e M - 1 , 0 u M
- 1 , 1 .times. e M - 1 , 0 u M - 1 , r - 1 .times. e M - 1 , 0 ] ,
##EQU00037##
where N.sub.h represents the quantity of the antennas in the first
dimension in the antenna group, N.sub.v represents the quantity of
the antennas in the second dimension in the antenna group, and r
represents the quantity of the data streams.
[0082] In a possible embodiment of the present disclosure, presumed
that e.sub.0,0=e.sub.1,0= . . . =e.sub.M-1,0 and an i.sup.th vector
in the column selection section is adopted, the precoding matrix W
may be calculated using the following equation:
W = 1 N h .times. N v r .function. [ u 0 , 0 .times. X 0 .function.
( : , i ) u 0 , 1 .times. X 0 .function. ( : , i ) u 0 , r - 1
.times. X 0 .function. ( : , i ) u 1 , 0 .times. X 1 .function. ( :
, i ) u 1 , 1 .times. X 1 .function. ( : , i ) u 1 , r - 1 .times.
X 1 .function. ( : , i ) u 2 , 0 .times. X 2 .function. ( : , i ) u
2 , 1 .times. X 2 .function. ( : , i ) u 2 , r - 1 .times. X 2
.function. ( : , i ) u M - 1 , 0 .times. X M - 1 .function. ( : , i
) u M - 1 , 1 .times. X M - 1 .function. ( : , i ) u M - 1 , r - 1
.times. X M - 1 .function. ( : , i ) ] , ##EQU00038##
where X.sub.i(:,i) represents the i.sup.th column of the matrix to
be selected.
[0083] When the quantity M of the antenna groups is M=4, the
quantity r of the data streams is r=3, X.sub.0=X.sub.1= . . .
=X.sub.M-1 and e.sub.0,0=e.sub.1,0= . . . =e.sub.M-1,0, the
codebook may be shown in Table 2.
TABLE-US-00002 TABLE 2 codebook where the quantity M of the antenna
groups is M = 4 and the quantity r of the data streams is r = 3
i.sub.2 i.sub.1,1,i.sub.1,2 0 1 2 3 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(3)
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(3)
w.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(3)
w.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(3) i.sub.1,2 = 0,
1, . . . , N.sub.vO.sub.2/2 - 1 i.sub.2 i.sub.1,1,i.sub.1,2 4 5 6 7
i.sub.1,1 = 0, 1, . . . , N.sub.hO.sub.1/2 - 1 {tilde over
(w)}.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(3) {tilde over
(w)}.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(3) {tilde over
(w)}.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(3)
w.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(3) i.sub.1,2 = 0,
1, . . . , N.sub.vO.sub.2/2 - 1 i.sub.2 i.sub.1,1,i.sub.1,2 8 9 10
11 i.sub.1,1 = 0, 1, . . . , N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(3)
w.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(3)
w.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(3)
w.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(3) i.sub.1,2 = 0,
1, . . . , N.sub.vO.sub.2/2 - 1 i.sub.2 i.sub.1,1,i.sub.1,2 12 13
14 15 i.sub.1,1 = 0, 1, . . . , N.sub.hO.sub.1/2 - 1
w.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(3)
w.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(3)
w.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(3)
w.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(3) i.sub.1,2 = 0,
1, . . . , N.sub.vO.sub.2/2 - 1 where W ( m 1 , m 2 ) ( 3 ) = 1 4
.times. N h .times. N v r [ v mod .function. ( m 2 , N v .times. O
2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m
2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 , N v
.times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - 1 v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) - 1 v mod .function. ( m 2 , N
v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - 1 v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) - 1 v mod .function. ( m 2 , N
v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) ] , ##EQU00039## W ~ ( m 1 , m 2 ) ( 3 ) = 1 4
.times. N h .times. N v r [ v mod .function. ( m 2 , N v .times. O
2 ) h mod .function. ( m 1 , N h .times. O 1 ) - j v mod .function.
( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1
) - 1 v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. (
m 1 , N h .times. O 1 ) j v mod .function. ( m 2 , N v .times. O 2
) h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2
, N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) j v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) - 1 v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 , N h .times. O 1 ) - j v mod .function. ( m 2
, N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) 1 v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) - j v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 , N
v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - j v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) ] ##EQU00040## W _ ( m 1 , m 2 ) ( 3 ) = 1 4
.times. N h .times. N v r [ v mod .function. ( m 2 , N v .times. O
2 ) h mod .function. ( m 1 , N h .times. O 1 ) - 1 v mod .function.
( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1
) v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1
, N h .times. O 1 ) - 1 v mod .function. ( m 2 , N v .times. O 2 )
h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 ,
N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 , N v
.times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) - 1 v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) - 1 v mod .function. ( m 2 , N
v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) ] ##EQU00041## W ^ ( m 1 , m 2 ) ( 3 ) = 1 4 .times.
N h .times. N v r [ v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 , N h .times. O 1 ) j v mod .function. ( m 2 ,
N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - 1 v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) - j v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 , N
v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - j v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) - 1 v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 , N h .times. O 1 ) j v mod .function. ( m 2 ,
N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) j v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 , N v
.times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) j v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) ] ##EQU00042##
TABLE-US-00003 TABLE 2-A codebook where the quantity M of the
antenna groups is M = 4 and the quantity r of the data streams is r
= 4 i.sub.2 i.sub.1,1,i.sub.1,2 0 1 2 3 4 5 6 7 i.sub.1,1 = 0, 1, .
. . , N.sub.hO.sub.1/2 - 1 i.sub.1,2 = 0, 1, . . . ,
N.sub.vO.sub.2/2 - 1 i.sub.2 i.sub.1,1,i.sub.1,2 8 9 10 11 12 13 14
15 i.sub.1,1 = 0, 1, . . . , N.sub.hO.sub.1/2 - 1 i.sub.1,2 = 0, 1,
. . . , N.sub.vO.sub.2/2 - 1 where W ( m 1 , m 2 , n ) ( 4 ) = 1 4
.times. N h .times. N v r Z n ( v mod .function. ( m 2 , N v
.times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) ) ,
##EQU00043## where Z.sub.n belongs to a set of matrices and each
matrix in the set having a dimension of 4 .times. 4 indicates data
missing or illegible when filed
[0084] Based on the above definition of the codebook,
(i.sub.1,1,i.sub.1,2) is used by the first-level codebook as a
first-level PMI1, and i.sub.2 is used by the second-level codebook
as a second-level PMI2. Different granularities of feedback
frequency domains and feedback periods may be adopted by the PMI1
and the PMI2. In a possible embodiment of the present disclosure,
the PMI1 is fed back in a long-time manner using a wide band, while
the PMI2 is fed back in a short-term manner using a subband. The
access device may determine the precoding matrix for the data
transmission in accordance with the feedback of the two levels of
codebooks.
Example 2
[0085] For the dual-polarized antenna array, r.sub.1=2, and the
antenna groups each include a same quantity of antenna elements,
i.e., N.sub.0=N.sub.1= . . . =N.sub.M-1.
[0086] For example, the relationship between the quantity of the
antenna groups of the antenna array and the rank r of the precoding
matrix is shown in Table 3.
TABLE-US-00004 TABLE 3 relationship between the quantity of antenna
groups of antenna array and rank of precoding matrix Rank .sup.r of
precoding The quantity .sup.M of the antenna matrix groups of
antenna array 1 2 2 2 3 2 4 2 5 4 6 4 7 4 8 4
[0087] In a possible embodiment of the present disclosure, the
beams for each antenna groups may be selected from a same set of
beam vectors, and the quantity L.sub.i of the beams included in the
beam group matrices X.sub.i may be the same.
[0088] W.sub.1 may be constructed as follows.
[0089] W.sub.1 may be constructed in a similar way as Example 1. As
mentioned above, in a possible embodiment of the present
disclosure, each beam group may include the beams orthogonal to
each other, so the beam group matrix X.sub.i may be slightly
different from that in Example 1. Every L.sub.i vectors in the set
of beam vectors may form one group, and L.sub.i=8. At this time, in
a possible grouping mode,
##EQU00044##
where i.sub.1,1=0, 1, . . . , N.sub.hO.sub.1/2-1, and i.sub.1,2=0,
1, . . . , N.sub.vO.sub.2/2-1. In this grouping mode, beams 1, 2,
3, 4 are orthogonal to beams 5, 6, 7 and 8 respectively. The entire
set of beam vectors may include
(N.sub.hO.sub.1/2)(N.sub.vO.sub.2/2) beam groups. Hence, with
respect to the beam group X.sub.i for each antenna group,
X.sub.i.di-elect cons.{X.sup.(i.sup.1,1.sup.,i.sup.1,2.sup.),
i.sub.1,1=0, 1, . . . , N.sub.hO.sub.1/2-1, i.sub.1,2=0, 1, . . . ,
N.sub.vO.sub.2/2-1}. The other procedures are the same as those
mentioned in Example 1, and thus will not be particularly defined
herein.
[0090] W.sub.2 may be constructed as follows. Because
r.sub.1=2,
W 2 ' = [ .times. [ e 0 , 0 , e 0 , 1 ] 0 [ e 1 , 0 , e 1 , 1 ] [ e
2 , 0 , e 2 , 1 ] 0 [ e M - 1 , 0 , e M - 1 , 1 ] ] ( U I 2 .times.
2 ) , ##EQU00045##
where e.sub.i,0 is a column selection vector having a length of 8
(because L.sub.i=8). When e.sub.i,0 is a unit vector, it may
include one element of 1 (indicating that the beams corresponding
to this column may be selected) and the other elements of 0. For
one antenna group, in order to ensure the selected two beams to be
orthogonal to each other, there are four possible column selection
sections, i.e.,
[ e i , 0 , e i , 1 ] .di-elect cons. { [ 1 0 0 0 0 0 0 0 0 1 0 0 0
0 0 0 ] , [ 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 ] , [ 0 0 0 0 1 0 0 0 0
0 0 0 0 1 0 0 ] , [ 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 ] } .
##EQU00046##
Hence, there exist 4.sup.M possible combinations for the column
selection sections of W.sub.2.
[0091] The phase adjustment matrix U among the antenna groups is an
M.times.r.sub.2 matrix, and it may be expressed as:
U = [ u 0 u 1 u r 2 - 1 ] = [ u 0 , 0 u 0 , 1 u 0 , r 2 - 1 u 1 , 0
u 1 , 1 u 1 , r 2 - 1 u M - 1 , 0 u M - 1 , 1 u M - 1 , r 2 - 1 ] .
##EQU00047##
Assume there are Q phase adjustment matrices, i.e., U.sub.0,
U.sub.1, U.sub.2, . . . , U.sub.Q-1. In this regard, the
second-level codebook may include 4.sup.MQ codewords, and i.sub.2
(i.sub.2=0, 1, . . . , 4.sup.MQ-1) may be adopted as a PMI to
indicate and feed back the precoding matrix. As mentioned above,
r.sub.2 is a minimum integer greater than or equal to r/r.sub.1
so
r 2 = r + 1 2 . ##EQU00048##
[0092] For the codebook where rank=4, M=2 as shown in Table 3. The
phase adjustment matrices U may be adopted to directly extend a
codebook of a 2-antenna LTE system (where rank=2) as follows:
{ 1 2 .function. [ 1 1 1 - 1 ] .times. .times. 1 2 .function. [ 1 1
e j .times. .pi. 4 - e j .times. .pi. 4 ] .times. .times. 1 2
.function. [ 1 1 j - j ] .times. .times. 1 2 .function. [ 1 1 e j
.times. 3 .times. .times. .pi. 4 - e j .times. 3 .times. .times.
.pi. 4 ] } . ##EQU00049##
Hence, there are Q=4 possible combinations for the phase adjustment
sections of W.sub.2.
[0093] The precoding matrix may be normalized so as to acquire the
final precoding matrix. When each antenna group is a
two-dimensional antenna array and the above phase adjustment matrix
among the antenna groups is adopted, the normalization operation
may be performed using the following equation:
W = 1 N h .times. N v r .function. [ X 0 0 X 1 .times. X 2 0 X M -
1 ] .times. col r ( [ u 0 , 0 .times. e 0 , 0 u 0 , 0 .times. e 0 ,
1 u 0 , 1 .times. e 0 , 0 u 0 , 1 .times. e 0 , 1 u 0 , r 2 - 1
.times. e 0 , 0 u 0 , r 2 - 1 .times. e 0 , 1 u 1 , 0 .times. e 1 ,
0 u 1 , 0 .times. e 1 , 1 u 1 , 1 .times. e 1 , 0 u 1 , 1 .times. e
1 , 1 u 1 , r 2 - 1 .times. e 1 , 0 u 1 , r 2 - 1 .times. e 1 , 1 u
2 , 0 .times. e 2 , 0 u 2 , 0 .times. e 2 , 1 u 2 , 1 .times. e 2 ,
0 u 2 , 1 .times. e 2 , 1 u 2 , r 2 - 1 .times. e 2 , 0 u 2 , r 2 -
1 .times. e 2 , 1 u M - 1 , 0 .times. e M - 1 , 0 u M - 1 , 0
.times. e M - 1 , 1 u M - 1 , 1 .times. e M - 1 , 0 u M - 1 , 1
.times. e M - 1 , 1 u M - 1 , r 2 - 1 .times. e M - 1 , 0 u M - 1 ,
r 2 - 1 e M - 1 , 1 ] ) , ##EQU00050##
where N.sub.h represents the quantity of the antennas in the first
dimension in the antenna group, N.sub.v represents the quantity of
the antennas in the second dimension in the antenna group, r
represents the quantity of the data streams, and col.sub.r( )
represents the r columns to be selected.
[0094] In a possible embodiment of the present disclosure, presumed
that [e.sub.0,0,e.sub.0,1]=[e.sub.1,0,e.sub.1,1]= . . .
[e.sub.M-1,0,e.sub.M-1,1] and an i.sup.th one in the column
selection combinations is adopted, the precoding matrix W may be
calculated using the following equation:
W = 1 N h .times. N v r .times. col r .function. ( [ u 0 , 0
.times. X 0 .function. ( : , i ) u 0 , 0 .times. X 0 .function. ( :
, i + 4 ) u 0 , r 2 - 1 .times. X 0 .function. ( : , i ) u 0 , r 2
- 1 .times. X 0 .function. ( : , i + 4 ) u 1 , 0 .times. X 1
.function. ( : , i ) u 1 , 0 .times. X 1 .function. ( : , i + 4 ) u
1 , r 2 - 1 .times. X 1 .function. ( : , i ) u 1 , r 2 - 1 .times.
X 0 .function. ( : , i + 4 ) u 2 , 0 .times. X 2 .function. ( : , i
) u 2 , 0 .times. X 2 .function. ( : , i + 4 ) u 2 , r 2 - 1
.times. X 2 .function. ( : , i ) u 2 , r 2 - 1 .times. X 0
.function. ( : , i + 4 ) u M - 1 , 0 .times. X M - 1 .function. ( :
, i ) u M - 1 , 0 .times. X M - 1 .function. ( : , i + 4 ) u M - 1
, r 2 - 1 .times. X M - 1 .function. ( : , i ) u M - 1 , r 2 - 1
.times. X M - 1 .function. ( : , i + 4 ) ] ) , ##EQU00051##
where X.sub.i(:,i) represents an i.sup.th column of the matrix to
be selected.
[0095] Presumed that X.sub.0=X.sub.1= . . . =X.sub.M-1 and
[e.sub.0,0,e.sub.0,1]=[e.sub.1,0,e.sub.1,1]= . . .
[e.sub.M-1,0,e.sub.M-1,1], the codebook where r.sub.1=2, M=2 and
rank=4 may be shown in Table 4.
TABLE-US-00005 TABLE 4 codebook where the quantity M of the antenna
groups is M = 2 and the quantity r of the data streams is r = 4
i.sub.2 i.sub.1,1,i.sub.1,2 0 1 2 3 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(4) i.sub.1,2 = 0,
1, . . . , N.sub.vO.sub.2/2 - 1 i.sub.2 i.sub.1,1,i.sub.1,2 4 5 6 7
i.sub.1,1 = 0, 1, . . . , N.sub.hO.sub.1/2 - 1 {tilde over
(W)}.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(4) {tilde over
(W)}.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(4) {tilde over
(W)}.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(4) {tilde over
(W)}.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(4) i.sub.1,2 =
0, 1, . . . , N.sub.vO.sub.2/2 - 1 i.sub.2 i.sub.1,1 8 9 10 11
i.sub.1,1 = 0, 1, . . . , N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(4) i.sub.1,2 = 0,
1, . . . , N.sub.vO.sub.2/2 - 1 i.sub.2 i.sub.1,1 12 13 14 15
i.sub.1,1 = 0, 1, . . . , N.sub.hO.sub.1/2 - 1
.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.).sup.(4)
.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.+1).sup.(4)
.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).sup.(4)
.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.+1).sup.(4) i.sub.1,2 = 0,
1, . . . , N.sub.vO.sub.2/2 - 1 where W ( m 1 , m 2 ) ( 4 ) = 1 2
.times. N h .times. N v r [ v mod .function. ( m 2 , N v .times. O
2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m
2 , N v .times. O 2 ) h mod .function. ( m 1 + O 1 , N h .times. O
1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m
1 , N h .times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 + O 1 , N h .times. O 1 ) v mod .function. ( m
2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 + O
1 , N h .times. O 1 ) - 1 v mod .function. ( m 2 , N v .times. O 2
) h mod .function. ( m 1 , N h .times. O 1 ) - 1 v mod .function. (
m 2 , N v .times. O 2 ) h mod .function. ( m 1 + O 1 , N h .times.
O 1 ) ] ##EQU00052## W ~ ( m 1 , m 2 ) ( 4 ) = 1 2 .times. N h
.times. N v r [ v mod .function. ( m 2 , N v .times. O 2 ) ? v mod
.function. ( m 2 , N v .times. O 1 ) h mod .function. ( m 1 + O 1 ,
N h .times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) ? h mod .function. ( m 1 + O 1
, N h .times. O 1 ) ? ? h mod .function. ( m 1 , N h .times. O 1 )
? v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1
+ O 1 , N h .times. O 1 ) - ? ? ? - ? ? h mod .function. ( m 1 + O
1 , N h .times. O 1 ) ] ##EQU00053## W _ ( m 1 , m 2 ) ( 4 ) = 1 2
.times. N h .times. N v r [ v mod .function. ( m 2 , N v .times. O
2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m
2 , N v .times. O 2 ) h mod .function. ( m 1 + O 1 , N h .times. O
1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m
1 , N h .times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 + O 1 , N h .times. O 1 ) j v mod .function. (
m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 )
j v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1
+ O 1 , N h .times. O 1 ) - j v mod .function. ( m 2 , N v .times.
O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - j v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 + O 1 ,
N h .times. O 1 ) ] ##EQU00054## W ^ ( m 1 , m 2 ) ( 4 ) = 1 2
.times. N h .times. N v r [ v mod .function. ( m 2 , N v .times. O
2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m
2 , N v .times. O 2 ) h mod .function. ( m 1 + O 1 , N h .times. O
1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m
1 , N h .times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 + O 1 , N h .times. O 1 ) ? v mod .function. (
m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 )
? v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1
+ O 1 , N h .times. O 1 ) - ? v mod .function. ( m 2 , N v .times.
O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - ? h mod
.function. ( m 1 + O 1 , N h .times. O 1 ) ] ##EQU00055##
Example 3
[0096] For the dual-polarized antenna array, when the r.sub.1 beams
selected from the beam group are not orthogonal to each other, in a
possible embodiment of the present disclosure, r.sub.1=2, and the
antenna groups each include a same quantity of antenna elements,
i.e., N.sub.0=N.sub.1= . . . =N.sub.M-1.
[0097] For example, the relationship between the quantity of the
antenna groups of the antenna array and the rank r of the precoding
matrix is shown in Table 5.
TABLE-US-00006 TABLE 5 relationship between the quantity of antenna
groups of antenna array and rank of precoding matrix Rank .sup.r of
precoding The quantity .sup.M of the antenna matrix groups of
antenna array 1 2 2 2 3 4 4 4 5 8 6 8 7 8 8 8
[0098] In this way, two antenna groups may form a superior group,
and there are
M 2 ##EQU00056##
superior groups. In a possible embodiment of the present
disclosure, the beams for each antenna groups may be selected from
a same set of beam vectors, and the quantity L.sub.i of the beams
included in the beam group matrices X.sub.i may be the same.
W 1 = [ X 0 0 X 1 X 2 0 X M - 1 ] ##EQU00057##
[0099] W.sub.1 may be constructed as follows:
[0100] W.sub.1 may be constructed in a same way as Example 1, which
will not be particularly defined herein. Because r.sub.1=2, every
two antenna groups may form one superior group. For example, every
two adjacent antenna groups may form one superior group, e.g.,
[X.sub.0,X.sub.1], [X.sub.2,X.sub.3], and the like. In a possible
embodiment of the present disclosure, the beam groups for the two
antenna groups in each superior group may be the same, i.e.,
X.sub.0=X.sub.1, X.sub.2=X.sub.3, . . . , X.sub.M-2=X.sub.M-1.
[0101] W.sub.2 may be constructed as follows. Because
r.sub.1=2,
W 2 ' = [ [ e 0 , 0 , e 0 , 1 ] 0 [ e 1 , 0 , e 1 , 1 ] [ e 2 , 0 ,
e 2 , 1 ] 0 [ e M - 1 , 0 , e M - 1 , 1 ] ] ( U V ) .
##EQU00058##
As mentioned in Example 1, e.sub.i,0 and e.sub.i,1 are column
selection vectors each having a length of 4 (because L.sub.i=4).
When each of e.sub.i,0 and e.sub.i,1 is a unit vector, it may
include one element of 1 (indicating that the beams corresponding
to this column may be selected) and the other elements of 0. For
one antenna group, it is unnecessary for the two column vectors to
be orthogonal to each other and two same beams may be selected, so
there are 10 possible column selection sections, i.e.,
[ e i , 0 , e i , 1 ] .di-elect cons. { [ 1 1 0 0 0 0 0 0 ] , [ 1 0
0 1 0 0 0 0 ] , [ 1 0 0 0 0 1 0 0 ] , [ 1 0 0 0 0 0 0 1 ] , [ 0 0 1
1 0 0 0 0 ] , [ 0 0 1 0 0 1 0 0 ] , .times. [ 0 0 1 0 0 0 0 1 ] , [
0 0 0 0 1 1 0 0 ] , [ 0 0 0 0 1 0 0 1 ] , [ 0 0 0 0 0 0 1 1 ] } .
##EQU00059##
[0102] In a possible embodiment of the present disclosure, the two
antenna groups in each superior group may select the same beams,
i.e., .left brkt-bot.e.sub.0,0,e.sub.0,1.right brkt-bot.=.left
brkt-bot.e.sub.1,0,e.sub.1,1.right brkt-bot., .left
brkt-bot.e.sub.2,0,e.sub.2,1.right brkt-bot.=.left
brkt-bot.e.sub.3,0,e.sub.3,1.right brkt-bot., . . . , .left
brkt-bot.e.sub.M=2,0,e.sub.M-2,1.right brkt-bot.=.left
brkt-bot.e.sub.M-1,0,e.sub.M-1,1.right brkt-bot..
[0103] The phase adjustment matrix V (V=[V'(:,1)e.sub.2.sup.(1)
V'(:,2)e.sub.2.sup.(2)]) in each superior group is a 4.times.2
matrix, where
V ' = [ .alpha. 0 .beta. 0 .alpha. 1 .beta. 1 ] , ##EQU00060##
i.e., a 2.times.2 orthogonal matrix which may be constructed using
the construction method mentioned in Example 1. For example, V' may
be constructed through Householder transform. A possible set of the
matrices V' may be
{ 1 2 .function. [ 1 1 1 - 1 ] .times. .times. 1 2 .function. [ 1 1
e j .times. .pi. 4 - e j .times. .pi. 4 ] .times. .times. 1 2
.function. [ 1 1 j - j ] .times. .times. 1 2 .function. [ 1 1 e j
.times. 3 .times. .times. .pi. 4 - e j .times. 3 .times. .times.
.pi. 4 ] } . ##EQU00061##
[0104] The phase adjustment matrix U among the superior groups is
a
M 2 .times. r 2 ##EQU00062##
orthogonal matrix, where r.sub.2 is a minimum integer greater than
or equal to
r 2 . ##EQU00063##
U may be expressed as:
U = [ u 0 , 0 u 0 , 1 u 0 , r 2 - 1 u 1 , 0 u 1 , 1 u 1 , r 2 - 1 u
M / 2 - 1 , 0 u M / 2 - 1 , 1 u M / 2 - 1 , r 2 - 1 ] .
##EQU00064##
It may also be constructed using the construction method mentioned
in Example 1. When r=8, U may be constructed through Householder
transform, and a possible set of the matrices U may be
{ 1 2 .function. [ 1 1 1 1 1 1 - 1 - 1 1 - 1 1 - 1 1 - 1 - 1 - 1 ]
.times. .times. 1 2 .function. [ 1 - j - 1 j j 1 j 1 - 1 - j 1 j -
j 1 - j 1 ] .times. .times. 1 2 .function. [ 1 - 1 1 - 1 1 1 - 1 -
1 1 1 1 1 1 - 1 - 1 - 1 ] .times. .times. 1 2 .function. [ - 1 j 1
- j - j 1 - j 1 1 j - 1 - j j 1 j 1 ] } . ##EQU00065##
[0105] Upon the acquisition of the first-level codebook and the
second-level codebook as mentioned above, one matrix in the
first-level codebook may be multiplied by one matrix in the
second-level codebook, and a resultant matrix may be normalized so
as to acquire the precoding matrix in the codebook. When each
antenna group is a two-dimensional antenna array, the following
normalization operation may be performed:
W = 1 N b .times. N v r .function. [ X 0 0 X 1 X 2 0 X M - 1 ]
.times. .times. col r .times. ( [ u 0 , 0 .times. .alpha. 0 .times.
e 0 , 0 u 0 , 0 .times. .beta. 0 .times. e 0 , 1 u 0 , 1 .times.
.alpha. 0 .times. e 0 , 0 u 0 , 1 .times. .beta. 0 .times. e 0 , 1
u 0 , r 2 - 1 .times. .alpha. 0 .times. e 0 , 0 u 0 , r 2 - 1
.times. .beta. 0 .times. e 0 , 1 u 0 , 0 .times. .alpha. 1 .times.
e 1 , 0 u 0 , 0 .times. .beta. 1 .times. e 1 , 1 u 0 , 1 .times.
.alpha. 1 .times. e 1 , 0 u 0 , 1 .times. .beta. 1 .times. e 1 , 1
u 0 , r 2 - 1 .times. .alpha. 1 .times. e 1 , 0 u 0 , r 2 - 1
.times. .beta. 1 .times. e 1 , 1 u 1 , 0 .times. .alpha. 0 .times.
e 2 , 0 u 1 , 0 .times. .beta. 0 .times. e 2 , 1 u 1 , 1 .times.
.alpha. 0 .times. e 2 , 0 u 1 , 1 .times. .beta. 0 .times. e 2 , 1
u 1 , r 2 - 1 .times. .alpha. 0 .times. e 2 , 0 u 1 , r 2 - 1
.times. .beta. 0 .times. e 2 , 1 u 1 , 0 .times. .alpha. 1 .times.
e 3 , 0 u 1 , 0 .times. .beta. 1 .times. e 3 , 1 u 1 , 1 .times.
.alpha. 1 .times. e 3 , 0 u 1 , 1 .times. .beta. 1 .times. e 3 , 1
u 1 , r 2 - 1 .times. .alpha. 1 .times. e 3 , 0 u 1 , r 2 - 1
.times. .beta. 1 .times. e 3 , 1 u Mi .times. .times. 2 - 1 , 0
.times. .alpha. 0 .times. e M - 2 , 0 u Mi .times. .times. 2 - 1 ,
0 .times. .beta. 0 .times. e M - 2 , 0 u Mi .times. .times. 2 - 1 ,
1 .times. .alpha. 0 .times. e M - 2 , 0 u Mi .times. .times. 2 - 1
, 1 .times. .beta. 0 .times. e M - 2 , 0 u Mi .times. .times. 2 - 1
, r 2 - 1 .times. .alpha. 0 .times. e M - 2 , 0 u Mi .times.
.times. 2 - 1 , r 2 - 1 .times. .beta. 0 .times. e M - 2 , 0 u Mi
.times. .times. 2 - 1 , 0 .times. .alpha. 1 .times. e M - 1 , 0 u
Mi .times. .times. 2 - 1 , 0 .times. .beta. 1 .times. e M - 1 , 0 u
Mi .times. .times. 2 - 1 , 1 .times. .alpha. 1 .times. e M - 1 , 0
u Mi .times. .times. 2 - 1 , 1 .times. .beta. 1 .times. e M - 1 , 0
u Mi .times. .times. 2 - 1 , r 2 - 1 .times. .alpha. 1 .times. e M
- 1 , 0 u Mi .times. .times. 2 - 1 , r 2 - 1 .times. .beta. 1
.times. e M - 1 , 0 ] ) , ##EQU00066##
where N.sub.h represents the quantity of the antennas in the first
dimension in the antenna group, N.sub.v represents the quantity of
the antennas in the second dimension in the antenna group, r
represents the quantity of the data streams, and col.sub.v( )
represents the r columns to be selected.
[0106] In a possible embodiment of the present disclosure, presumed
that [e.sub.0,0,e.sub.0,1]=[e.sub.1,0,e.sub.1,1]= . . .
[e.sub.M-1,0,e.sub.M-1,1] and an i.sup.th vector and a j.sup.th
vector in the four column selection combinations are adopted, the
precoding matrix W may be calculated using the following
equation:
W = 1 N h .times. N v r .times. .times. col r .times. ( [ u 0 , 0
.times. .alpha. 0 .times. X 0 .function. ( : , j ) u 0 , 0 .times.
.beta. 0 .times. X 0 .function. ( : , j ) u 0 , 1 .times. .alpha. 0
.times. X 0 .function. ( : , j ) u 0 , 1 .times. .beta. 0 .times. X
0 .function. ( : , j ) u 0 , 0 .times. .alpha. 1 .times. X 1
.function. ( : , j ) u 0 , 0 .times. .beta. 1 .times. X 1
.function. ( : , j ) u 0 , 1 .times. .alpha. 1 .times. X 1
.function. ( : , j ) u 0 , 1 .times. .beta. 1 .times. X 1
.function. ( : , j ) u 1 , 0 .times. .alpha. 0 .times. X 2
.function. ( : , j ) u 1 , 0 .times. .beta. 0 .times. X 2
.function. ( : , j ) u 1 , 1 .times. .alpha. 0 .times. X 2
.function. ( : , j ) u 1 , 1 .times. .beta. 0 .times. X 2
.function. ( : , j ) u 1 , 0 .times. .alpha. 1 .times. X 3
.function. ( : , j ) u 1 , 0 .times. .beta. 1 .times. X 3
.function. ( : , j ) u 1 , 1 .times. .alpha. 1 .times. X 3
.function. ( : , j ) u 1 , 1 .times. .beta. 1 .times. X 3
.function. ( : , j ) u Mi .times. .times. 2 - 1 , 0 .times. .alpha.
0 .times. X M - 2 .function. ( : , j ) u Mi .times. .times. 2 - 1 ,
0 .times. .beta. 0 .times. X M - 2 .function. ( : , j ) u Mi
.times. .times. 2 - 1 , 1 .times. .alpha. 0 .times. X M - 2
.function. ( : , j ) u Mi .times. .times. 2 - 1 , 1 .times. .beta.
0 .times. X M - 2 .function. ( : , j ) u Mi .times. .times. 2 - 1 ,
0 .times. .alpha. 1 .times. X M - 1 .function. ( : , j ) u Mi
.times. .times. 2 - 1 , 0 .times. .beta. 1 .times. X M - 1
.function. ( : , j ) u Mi .times. .times. 2 - 1 , 1 .times. .alpha.
1 .times. X M - 1 .function. ( : , j ) u Mi .times. .times. 2 - 1 ,
1 .times. .beta. 1 .times. X M - 1 .function. ( : , j ) .times.
.times. .times. u 0 , r 2 - 1 .times. .alpha. 0 .times. X 0
.function. ( : , j ) u 0 , r 2 - 1 .times. .beta. 0 .times. X 0
.function. ( : , j ) u 0 , r 2 - 1 .times. .alpha. 1 .times. X 1
.function. ( : , j ) u 0 , r 2 - 1 .times. .beta. 1 .times. X 1
.function. ( : , j ) u 1 , r 2 - 1 .times. .alpha. 0 .times. X 2
.function. ( : , j ) u 1 , r 2 - 1 .times. .beta. 0 .times. X 2
.function. ( : , j ) u 1 , r 2 - 1 .times. .alpha. 1 .times. X 3
.function. ( : , j ) u 1 , r 2 - 1 .times. .beta. 1 .times. X 3
.function. ( : , j ) u Mi .times. .times. 2 - 1 , r 2 - 1 .times.
.alpha. 0 .times. X M - 2 .function. ( : , j ) u Mi .times. .times.
2 - 1 , r 2 - 1 .times. .beta. 0 .times. X M - 2 .function. ( : , j
) u Mi .times. .times. 2 - 1 , r 2 - 1 .times. .alpha. 1 .times. X
M - 1 .function. ( : , j ) u Mi .times. .times. 2 - 1 , r 2 - 1
.times. .beta. 1 .times. X M - 1 .function. ( : , j ) ] ) ,
##EQU00067##
where X.sub.i(:,i) represents an i.sup.th column of the matrix to
be selected.
[0107] Presumed that X.sub.0=X.sub.1= . . . =X.sub.M-1 each beam
group for each antenna group includes four beams and
[e.sub.0,0,e.sub.0,1]=[e.sub.1,0,e.sub.1,1]= . . .
[e.sub.M-1,0,e.sub.M-1,1], the codebook where r.sub.1=2, M=4 and
rank=4 may be shown in Table 6.
TABLE-US-00007 TABLE 6 codebook where the quantity M of the antenna
groups is M = 4 and the quantity r of the data streams is r = 4
i.sub.2 i.sub.1,1,i.sub.1,2 0 1 2 3 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.,2i.sub.1,2.sub.).s-
up.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+1,2i.sub.1-
,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.+2,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.sub.1,2.sub-
.).sup.(4)
W.sub.(2i.sub.1,1.sub.+3,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.s-
ub.1,2.sub.).sup.(4) i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
i.sub.2 i.sub.1,1 4 5 6 7 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+1,2i.sub.1,2.sub.)-
.sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.sub-
.1,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub.1,2.sub.)-
.sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub-
.1,2.sub.).sup.(4) i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
i.sub.2 i.sub.1,1,i.sub.1,2 8 9 10 11 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1 i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
i.sub.2 i.sub.1,1 12 13 14 15 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1 {tilde over
(W)}.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+1,2i.sub.1,2.su-
b.).sup.(4) {tilde over
(W)}.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.sub.1,2.-
sub.).sup.(4) {tilde over
(W)}.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub.1,2.su-
b.).sup.(4) {tilde over
(W)}.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub.1,2.-
sub.).sup.(4) i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
i.sub.2 i.sub.1,1,i.sub.1,2 16 17 18 19 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.,2i.sub.1,2.sub.).s-
up.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+1,2i.sub.1-
,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.+2,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.sub.1,2.sub-
.).sup.(4)
W.sub.(2i.sub.1,1.sub.+3,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.s-
ub.1,2.sub.).sup.(4) i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
i.sub.2 i.sub.1,1 20 21 22 23 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+1,2i.sub.1,2.sub.)-
.sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.sub-
.1,2.sub.).sup.(4)
W.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub.1,2.sub.)-
.sup.(4)
W.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub-
.1,2.sub.).sup.(4) i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
i.sub.2 i.sub.1,1,i.sub.1,2 24 25 26 27 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.,2i.sub.1,2.sub.).su-
p.(4)
.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+1,2i.sub.1,-
2.sub.).sup.(4)
.sub.(2i.sub.1,1.sub.+2,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.sub.1,2.sub.-
).sup.(4)
.sub.(2i.sub.1,1.sub.+3,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.su-
b.1,2.sub.).sup.(4) i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
i.sub.2 i.sub.1,1 28 29 30 11 i.sub.1,1 = 0, 1, . . . ,
N.sub.hO.sub.1/2 - 1
.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+1,2i.sub.1,2.sub.).-
sup.(4)
.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+2,2i.sub.-
1,2.sub.).sup.(4)
.sub.(2i.sub.1,1.sub.,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub.1,2.sub.).-
sup.(4)
.sub.(2i.sub.1,1.sub.+1,2i.sub.1,2.sub.,2i.sub.1,1.sub.+3,2i.sub.-
1,2.sub.).sup.(4) i.sub.1,2 = 0, 1, . . . , N.sub.vO.sub.2/2 - 1
where W ( m 1 , m 2 , m 1 ' , m 2 ' ) ( 4 ) = 1 4 .times. N h
.times. N v r .times. ? [ v mod .function. ( m 2 , N v .times. O 2
) ? h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m
2 ' , N v .times. O 2 ) ? ? ? v mod .function. ( m 2 ' , N v
.times. O 2 ) ? v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 ' , N v
.times. O 2 ) ? ? h mod .function. ( m 1 , N h .times. O 1 ) - 1 ?
? ? .times. v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) ? ? - ? v mod .function. ( m 2
, N v .times. O 2 ) ? - ? v mod .function. ( m 2 ' , N v .times. O
2 ) ? ? h mod .function. ( m 1 , N h .times. O 1 ) - ? .times. v
mod .function. ( m 2 ' , N v .times. O 2 ) ? - ? v mod .function. (
m 2 , N v .times. O 2 ) ? ? v mod .function. ( m 2 ' , N v .times.
O 2 ) ? ] ##EQU00068## W ~ ( m 1 , m 2 , m 1 ' , m 2 ' ) ( 4 ) = 1
4 .times. N h .times. N v r [ v mod .function. ( m 2 , N v .times.
O 2 ) ? h mod .function. ( m 1 , N h .times. O 1 ) v mod .function.
( m 2 ' , N v .times. O 2 ) ? v mod .function. ( m 2 , N v .times.
O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod .function. (
m 2 ' , N v .times. O 2 ) h mod .function. ( m 1 ' , N h .times. O
1 ) v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m
1 , N h .times. O 1 ) - 1 v mod .function. ( m 2 ' , N v .times. O
2 ) ? v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. (
m 1 , N h .times. O 1 ) - 1 v mod .function. ( m 2 ' , N v .times.
O 2 ) h mod .function. ( m 1 ' , N h .times. O 1 ) ? v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) ? v mod .function. ( m 2 ' , N v .times. O 2 ) ? - ?
v mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 ,
N h .times. O 1 ) - ? v mod .function. ( m 2 ' , N v .times. O 2 )
h mod .function. ( m 1 ' , N h .times. O 1 ) ? .times. v mod
.function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h
.times. O 1 ) - ? ? - ? v mod .function. ( m 2 , N v .times. O 2 )
h mod .function. ( m 1 , N h .times. O 1 ) ? h mod .function. ( m 1
' , N h .times. O 1 ) ] ##EQU00069## W _ ( m 1 , m 2 , m 1 ' , m 2
' ) ( 4 ) = 1 4 .times. N h .times. N v r [ v mod .function. ( m 2
, N v .times. O 2 ) ? h mod .function. ( m 1 , N h .times. O 1 ) v
mod .function. ( m 2 ' , N v .times. O 2 ) h mod .function. ( m 1 '
, N h .times. O 1 ) v mod .function. ( m 2 , N v .times. O 2 ) ? h
mod .function. ( m 1 , N h .times. O 1 ) v mod .function. ( m 2 ' ,
N v .times. O 2 ) h mod .function. ( m 1 ' , N h .times. O 1 ) j v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) - j v mod .function. ( m 2 ' , N v .times. O 2 ) h
mod .function. ( m 1 ' , N h .times. O 1 ) j v mod .function. ( m 2
, N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) - j
v mod .function. ( m 2 ' , N v .times. O 2 ) h mod .function. ( m 1
' , N h .times. O 1 ) ? .times. v mod .function. ( m 2 , N v
.times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) ? h mod
.function. ( m 1 ' , N h .times. O 1 ) - ? .times. v mod .function.
( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1
) - ? v mod .function. ( m 2 ' , N v .times. O 2 ) h mod .function.
( m 1 ' , N h .times. O 1 ) ? h mod .function. ( m 1 , N h .times.
O 1 ) - ? .times. v mod .function. ( m 2 ' , N v .times. O 2 ) h
mod .function. ( m 1 ' , N h .times. O 1 ) - ? v mod .function. ( m
2 , N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) ?
h mod .function. ( m 1 ' , N h .times. O 1 ) ] ##EQU00070## W ^ ( m
1 , m 2 , m 1 ' , m 2 ' ) ( 4 ) = 1 4 .times. N h .times. N v r [ v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) v mod .function. ( m 2 ' , N v .times. O 2 ) h mod
.function. ( m 1 ' , N h .times. O 1 ) v mod .function. ( m 2 , N v
.times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) v mod
.function. ( m 2 ' , N v .times. O 2 ) h mod .function. ( m 1 ' , N
h .times. O 1 ) j v mod .function. ( m 2 , N v .times. O 2 ) h mod
.function. ( m 1 , N h .times. O 1 ) - j v mod .function. ( m 2 ' ,
N v .times. O 2 ) h mod .function. ( m 1 ' , N h .times. O 1 ) j v
mod .function. ( m 2 , N v .times. O 2 ) h mod .function. ( m 1 , N
h .times. O 1 ) - j v mod .function. ( m 2 ' , N v .times. O 2 ) h
mod .function. ( m 1 ' , N h .times. O 1 ) ? v mod .function. ( m 2
, N v .times. O 2 ) h mod .function. ( m 1 , N h .times. O 1 ) ? v
mod .function. ( m 2 ' , N v .times. O 2 ) h mod .function. ( m 1 '
, N h .times. O 1 ) - ? v mod .function. ( m 2 , N v .times. O 2 )
h mod .function. ( m 1 , N h .times. O 1 ) - ? v mod .function. ( m
2 ' , N v .times. O 2 ) h mod .function. ( m 1 ' , N h .times. O 1
) ? h mod .function. ( m 1 , N h .times. O 1 ) - ? .times. v mod
.function. ( m 2 ' , N v .times. O 2 ) h mod .function. ( m 1 ' , N
h .times. O 1 ) - ? v mod .function. ( m 2 , N v .times. O 2 ) h
mod .function. ( m 1 , N h .times. O 1 ) ? h mod .function. ( m 1 '
, N h .times. O 1 ) ] ##EQU00071## indicates data missing or
illegible when filed
[0108] The phase adjustment matrix V' may be
V ' .di-elect cons. { 1 2 .function. [ 1 1 1 - 1 ] .times. .times.
1 2 .function. [ 1 1 j - j ] } , ##EQU00072##
and the phase adjustment matrix U may be
U .di-elect cons. { 1 2 .function. [ 1 1 e j .times. .pi. 4 - e j
.times. .pi. 4 ] .times. .times. 1 2 .function. [ 1 1 e j .times. 3
.times. .pi. 4 - e j .times. 3 .times. .pi. 4 ] } .
##EQU00073##
Based on the above definition of the codebook,
(i.sub.1,1,i.sub.1,2) is used by the first-level codebook as a
first-level PMI1, and i.sub.2 is used by the second-level codebook
as a second-level PMI2. Different granularities of feedback
frequency domains and feedback periods may be adopted by the PMI1
and the PMI2. In a possible embodiment of the present disclosure,
the PMI1 is fed back in along-time manner using a wide band, while
the PMI2 is fed back in a short-term manner using a subband. The
base station may determine the precoding matrix for the data
transmission in accordance with the feedback of the two levels of
codebooks.
Second Embodiment
[0109] As shown in FIG. 5, the present disclosure further provides
in this embodiment a CSI feedback method implemented by an access
device, which includes: Step S501 of receiving a first precoding
matrix index value and a second precoding matrix index value from a
UE, the first precoding matrix index value corresponding to a first
precoding matrix W.sub.1 in a set of first precoding matrices
determined by the UE, the first precoding matrix W.sub.1 consisting
of beam groups for a plurality of antenna groups of an antenna
array; the second precoding matrix index value corresponding to a
second precoding matrix W.sub.2 in a set of second precoding
matrices determined by the UE, the second precoding matrix W.sub.2
including a beam selection section for selecting beams from the
beam groups for each antenna group and a phase adjustment section
for adjusting phases among the plurality of antenna groups; and
Step S502 of selecting corresponding precoding matrices from the
set of first precoding matrices and the set of second precoding
matrices in accordance with the first precoding matrix index value
and the second precoding matrix index value, and implementing
calculation on the selected precoding matrices to obtain a
precoding matrix W for data transmission, where
W=W.sub.1W.sub.2.
[0110] In a possible embodiment of the present disclosure, the CSI
feedback method further includes notifying the UE of codebook
parameters of the codebook. The codebook parameters include one or
more of the quantities: quantity M of the antenna groups of the
antenna array, a grouping mode of the antenna groups of the antenna
array, the quantity N.sub.i of antenna elements in each antenna
group, the quantity of rows and columns of the antenna elements in
each antenna group, a sampling rate of the beams in a precoding
matrix, and a constitution mode of the beam groups for each antenna
group in W.sub.1.
[0111] In a possible embodiment of the present disclosure, the CSI
feedback method further includes: constructing a set of the first
precoding matrices W.sub.1 to obtain a first codebook and
constructing a set of the second precoding matrices W.sub.2 to
obtain a second codebook; and acquiring a codebook in accordance
with the matrices in the first codebook and the second codebook
through calculation.
[0112] In a possible embodiment of the present disclosure, the
first codebook is for describing a long-term characteristic of a
channel, so it may be fed back using a wide band in a long-time
manner. The second codebook is for describing a short-term
characteristic of the channel, so it may be fed back using a
subband in a short-term manner.
[0113] According to the CSI feedback method in the embodiment of
the present disclosure, the antenna array is divided into a
plurality of antenna groups. Through two levels of codebooks, it is
able to achieve the beam selection in each antenna group and the
phase adjustment among the antenna groups. The two levels of
codebooks are capable of being fed back respectively. As a result,
it is able to quantize the channel in a more accurate manner, and
improve the system performance. In addition, the quantity of the
antenna groups may vary along with the quantity of data
streams.
Third Embodiment
[0114] As shown in FIG. 6, the present disclosure further provides
in this embodiment a CSI feedback device 60, which includes: a
first determination module 601 configured to determine a first
precoding matrix W.sub.1 in a set of first precoding matrices, and
determine a first precoding matrix index value corresponding to the
first precoding matrix W.sub.1, the first precoding matrix W.sub.1
consisting of beam groups for a plurality of antenna groups of an
antenna array; a second predetermined module 602 configured to
determine a second precoding matrix W.sub.2 in a set of second
precoding matrices, and determine a second precoding matrix index
value corresponding to the second precoding matrix W.sub.2, the
second precoding matrix W.sub.2 including a beam selection section
for selecting beams from the beam groups for each antenna group and
a phase adjustment section for adjusting phases among the plurality
of antenna groups; and a first feedback module 603 configured to
transmit the first precoding matrix index value and the second
precoding matrix index value to an access device, so as to enable
the access device to select corresponding precoding matrices from
the set of first precoding matrices and the set of second precoding
matrix in accordance with the first precoding matrix index value
and the second precoding matrix index value, and calculate a
precoding matrix W for data transmission in accordance with the
selected precoding matrices, where W=W.sub.1W.sub.2.
[0115] In a possible embodiment of the present disclosure, the
precoding matrix W may be calculated using the following
equation:
W = W 1 .times. W 2 = [ X 0 0 X 1 X 2 0 X M - 1 ] W 2 .
##EQU00074##
The first precoding matrix W.sub.1 is a block diagonal matrix
including elements of 0 other than diagonal blocks where a beam
group matrix X.sub.i is located, i=0, 1, . . . , M-1, M represents
the quantity of the antenna groups for the antenna array, X.sub.i
represents a beam group for an i.sup.th antenna group and it is a
N.sub.i.times.L.sub.i matrix with each column representing one beam
direction, N.sub.i represents the quantity of antenna elements in
the i.sup.th antenna group, N represents a total quantity of
antenna elements for the antenna array and
N = i = 0 M - 1 .times. N i , ##EQU00075##
and L.sub.i represents the quantity of beams in the beam group for
the i.sup.th antenna group. The second precoding matrix W.sub.2 is
configured for beam selection and phase adjustment, and it is
capable of selecting several beams from each beam group and
adjusting a phase among the selected beams for each antenna group,
so as to acquire a final precoding matrix. The second precoding
matrix W.sub.2 is a
( i = 0 M - 1 .times. L i ) .times. r ##EQU00076##
matrix, where r represents a rank of the precoding matrix W.
[0116] In a possible embodiment of the present disclosure,
W 2 ' = [ E 0 0 E 1 E 2 0 E M - 1 ] W p , .times. where .times. [ E
0 0 E 1 E 2 0 E M - 1 ] ##EQU00077##
represents the beam selection section and used to select r.sub.1
beams from each antenna group, a beam selection vector for the
i.sup.th antenna group is defined as E.sub.i=.left
brkt-bot.e.sub.i,0 e.sub.i,1 . . . e.sub.i,r.sub.1.sub.-1.right
brkt-bot., where e.sub.i,j represents a column selection vector
having a length of L.sub.i and it is configured to select one beam
from the beam group matrix X.sub.i or merge the beams in X.sub.i
into one beam, j=0, 1, . . . r.sub.1-1, W.sub.p represents the
phase adjustment section, and r columns of W.sub.2' are selected as
the second precoding matrix W.sub.2.
[0117] In a possible embodiment of the present disclosure, when the
r.sub.1 beams are orthogonal to each other, W.sub.p=UI.sub.r.sub.1,
where U represents an orthogonal matrix including r.sub.2 columns,
each column includes M phase adjustment factors, r.sub.2 is a
minimum integer greater than or equal to
r r 1 , ##EQU00078##
I.sub.r.sub.1 represents a r.sub.1.times.r.sub.1 unit matrix. When
the r.sub.1 beams are not orthogonal to each other, every r groups
of M antenna groups form
M r 1 ##EQU00079##
superior groups, and W.sub.p=UV, where U represents an orthogonal
matrix for adjusting a phase among the superior groups and
including r.sub.2 columns, each column includes
M r 1 ##EQU00080##
phase adjustment factors, r.sub.2 is a minimum integer greater than
or equal to
r r 1 , ##EQU00081##
V is used for adjusting a phase among the antenna groups in one
superior group and V=.left brkt-bot.V'(:,1)e.sub.r.sub.1.sup.(1)
V'(:,2)e.sub.r.sub.1.sup.(2) . . .
V'(:,r.sub.1)e.sub.r.sub.1.sup.(r.sup.1.sup.).right brkt-bot., V'
represents an orthogonal matrix including r.sub.1 columns, each
column includes r.sub.1 phase adjustment factors, V'(:,i)
represents an i.sup.th column of V', and e.sub.r.sub.1.sup.(i)
represents an i.sup.th column of a unit matrix
I.sub.r.sub.1.sub..times.r.sub.1.
[0118] In a possible embodiment of the present disclosure, the
quantity M of the antenna groups of the antenna array varies along
with the rank r of the preboding matrix W.
[0119] In a possible embodiment of the present disclosure, the CSI
feedback device further includes a first codebook parameter
processing module configured to acquire a channel measurement
result, determine codebook parameters of a codebook in accordance
with the channel measurement result, and transmit the codebook
parameters to the access device, or determine one or more
parameters corresponding to the codebook parameters in accordance
with a predetermined mapping relationship between the codebook
parameters and the one or more parameters and transmit the one or
more parameters to the access device. A set of the first precoding
matrices W.sub.1 form a first codebook, a set of the second
precoding matrices W.sub.2 form a second codebook, and a set of the
precoding matrices for data transmission acquired after the
calculation form the codebook.
[0120] In a possible embodiment of the present disclosure, the CST
feedback device further includes a second codebook parameter
processing module configured to receive one or more parameters from
the access device, and determine codebook parameters corresponding
to the one or more parameters in accordance with a predetermined
mapping relationship between the codebook parameters and the one or
more parameters or receive all the codebook parameters from the
access device.
[0121] In a possible embodiment of the present disclosure, the
codebook parameters include one or more of the quantities: quantity
M of the antenna groups of the antenna array, a grouping mode of
the antenna groups of the antenna array, the quantity N.sub.i of
the antenna elements in each antenna group, the quantity of rows
and columns of the antenna elements in each antenna group, a
sampling rate of the beams in a precoding matrix, and a
constitution mode of the beam groups for each antenna group in
W.sub.1 (a beam configuration mode).
[0122] According to the CSI feedback device in the embodiment of
the present disclosure, the antenna array is divided into a
plurality of antenna groups. Through two levels of codebooks, it is
able to achieve the beam selection in each antenna group and the
phase adjustment among the antenna groups. The two levels of
codebooks are capable of being fed back respectively. As a result,
it is able to quantize the channel in a more accurate manner, and
improve the system performance. In addition, the quantity of the
antenna groups may vary along with the quantity of data
streams.
Fourth Embodiment
[0123] As shown in FIG. 7, the present disclosure further provides
in this embodiment a CSI feedback device 70, which includes: a
second feedback module 701 configured to receive a first precoding
matrix index value and a second precoding matrix index value from a
UE, the first precoding matrix index value corresponding to a first
precoding matrix W.sub.1 in a set of first precoding matrices
determined by the UE, the first precoding matrix W.sub.1 consisting
of beam groups for a plurality of antenna groups of an antenna
array, the second precoding matrix index value corresponding to a
second precoding matrix W.sub.2 in a set of second precoding
matrices determined by the UE, the second precoding matrix W.sub.2
including a beam selection section for selecting beams from the
beam groups for each antenna group and a phase adjustment section
for adjusting phases among the plurality of antenna groups; and a
precoding matrix selection module 702 configured to select
corresponding precoding matrices from the set of first precoding
matrices and the set of second precoding matrices in accordance
with the first precoding matrix index value and the second
precoding matrix index value, and calculate a precoding matrix W
for data transmission in accordance with the selected precoding
matrices, where W=W.sub.1W.sub.2.
[0124] In a possible embodiment of the present disclosure, the CSI
feedback device further includes a third feedback module configured
to notify the UE of codebook parameters of a codebook. The codebook
parameters include one or more of the quantities: quantity M of the
antenna groups of the antenna array, a grouping mode of the antenna
groups of the antenna array, the quantity N.sub.i of antenna
elements in each antenna group, the quantity of rows and columns of
the antenna elements in each antenna group, a sampling rate of the
beams in a precoding matrix, and a constitution mode of the beam
groups for each antenna group in W.sub.1. A set of the first
precoding matrices W.sub.1 form a first codebook, a set of the
second precoding matrices W.sub.2 form a second codebook, and a set
of the precoding matrices for data transmission acquired after the
calculation form the codebook.
[0125] According to the CSI feedback device in the embodiment of
the present disclosure, the antenna array is divided into a
plurality of antenna groups. Through two levels of codebooks, it is
able to achieve the beam selection in each antenna group and the
phase adjustment among the antenna groups. The two levels of
codebooks are capable of being fed back respectively. As a result,
it is able to quantize the channel in a more accurate manner, and
improve the system performance. In addition, the quantity of the
antenna groups may vary along with the quantity of data
streams.
[0126] It should be appreciated that, such phrases as "one
embodiment" or "one of the embodiments" intend to indicate that the
specific features, structures or characteristics are contained in
at least one embodiment the present disclosure, rather than
referring to a same embodiment. In addition, the features,
structures or characteristics may be combined in any embodiment or
embodiments in an appropriate manner.
[0127] In the embodiments of the present disclosure, the order of
the steps may not be limited to the serial numbers thereof, and it
shall be determined on the basis of the functions and internal
logic.
[0128] It should be further appreciated that, the device and method
may be implemented in any other ways. For example, the embodiments
for the apparatus is merely for illustrative purposes, and the
modules or units are provided merely on the basis of their logic
functions. During the actual application, some modules or units may
be combined together or integrated into another system.
Alternatively, some functions of the module or units may be omitted
or not executed. In addition, the coupling connection, direct
coupling connection or communication connection between the modules
or units may be implemented via interfaces, and the indirect
coupling connection or communication connection between the modules
or units may be implemented in an electrical or mechanical form or
in any other form.
[0129] In addition, the functional units in the embodiments of the
present disclosure may be integrated into a processing unit, or the
functional units may exist independently, or two or more functional
units may be combined together. These units may be implemented in
the form of hardware, or in the form of hardware plus software
functional units.
[0130] The software functional units may be stored in a
computer-readable storage medium, and may include several
instructions so as to enable a computer device (a personal
computer, a server or a network device) to execute parts of the
steps of the method mentioned in the embodiments of the present
disclosure. The storage medium may include any medium capable of
storing therein program codes, e.g., a universal serial bus (USB)
flash disk, a mobile hard disk (HD), a read-only memory (ROM), a
random access memory (RAM), a magnetic disk or an optical disk.
[0131] The above embodiments are for illustrative purposes only,
but the present disclosure is not limited thereto. Obviously, a
person skilled in the art may make further modifications and
improvements without departing from the spirit of the present
disclosure, and these modifications and improvements shall also
fall within the scope of the present disclosure.
* * * * *